WO2017064914A1

WO2017064914A1 - Information-processing device

Info

Publication number: WO2017064914A1
Application number: PCT/JP2016/073655
Authority: WO
Inventors: 俊之関矢; 裕一郎小山; 雄哉平野
Original assignee: ソニー株式会社
Priority date: 2015-10-13
Filing date: 2016-08-10
Publication date: 2017-04-20

Abstract

[Problem] To collect desired sounds in a more suitable manner even under an environment in which noise occurs at random. [Solution] An information-processing device provided with a sound-collecting unit and a support member provided in a portion at least with a convex part having a streamlined shape, the support member supporting the sound-collecting unit so that the sound-collecting unit is located at or near the tip of the convex part.

Description

Information processing device

This disclosure relates to an information processing apparatus.

In recent years, with the advancement of communication technology and the miniaturization of various devices, the type of equipment called so-called information processing devices has also diversified, not limited to PCs (Personal Computers) etc., like smartphones and tablet terminals, Information processing apparatuses configured to be carried by users are also becoming popular. In particular, in recent years, a so-called wearable device has also been proposed that is configured to be used while being carried by a user wearing it on a part of the body.

In recent years, with the development of so-called speech recognition technology and natural language processing technology, information processing apparatuses having a user interface (UI) that allows a user to instruct various processes by speech input have become widespread. ing.

JP 2012-203122 A

In addition, in an information processing apparatus configured to be able to collect a voice uttered by a user for voice recognition, a voice call, and the like, by suppressing other sounds (that is, noise) other than the voice to be collected. Therefore, a mechanism capable of further improving the sound collection quality of the sound has been studied. For example, Patent Document 1 discloses an example of a mechanism for suppressing noise.

On the other hand, it is assumed that the environment surrounding the information processing device changes dynamically as the usage scene of the information processing device diversifies, such as when the information processing device is used outdoors. Is done. Under such circumstances, it may be assumed that sound generated on the information processing apparatus such as wind noise or sound accompanying vibration is collected as noise. Such a sound becomes a noise which is randomly generated with irregular locations and times.

Therefore, the present disclosure proposes an information processing apparatus capable of collecting a target sound in a more preferable manner even in an environment where noise is randomly generated.

According to the present disclosure, at least a part of the sound collecting part and a convex part having a streamline shape is provided, and the sound collecting part is supported so as to be positioned at or near the tip of the convex part. An information processing apparatus comprising a support member is provided.

As described above, according to the present disclosure, there is provided an information processing apparatus that can collect a target sound in a more preferable manner even in an environment where noise is randomly generated.

Note that the above effects are not necessarily limited, and any of the effects shown in the present specification, or other effects that can be grasped from the present specification, together with or in place of the above effects. May be played.

4 is an explanatory diagram for describing an example of a schematic configuration of an information processing device according to a first embodiment of the present disclosure. FIG. Explanatory drawing for demonstrating an example of schematic structure of the information processing apparatus which concerns on the embodiment It is explanatory drawing for demonstrating an example of the observation environment for observing the influence of a wind noise. An example of an installation position of each of a plurality of sound collection units provided in the information processing apparatus is illustrated. It is explanatory drawing for demonstrating an example of the observation result of the wind noise by each said sound collection part when a wind is applied to an information processing apparatus from a mutually different angle. It is the block diagram which showed an example of the function structure of the information processing apparatus which concerns on the embodiment. The information processing apparatus which concerns on the embodiment has shown an example of the process which acquires a target sound based on the sound collection result of each of several sound collection parts. 5 is a flowchart illustrating an example of a flow of a series of processes of the information processing apparatus according to the embodiment. 1 is an explanatory diagram for describing an example of an information processing apparatus according to a first embodiment; FIG. 6 is an explanatory diagram for explaining another example of the information processing apparatus according to the first embodiment. FIG. 6 is an explanatory diagram for explaining another example of the information processing apparatus according to the first embodiment. FIG. 9 is an explanatory diagram for describing an example of an information processing apparatus according to a second embodiment. FIG. 10 is an explanatory diagram for explaining another example of the information processing apparatus according to the second embodiment. FIG. 10 is an explanatory diagram for explaining another example of the information processing apparatus according to the second embodiment. FIG. 10 is an explanatory diagram for explaining another example of the information processing apparatus according to the second embodiment. FIG. 9 is an explanatory diagram for describing an example of an information processing apparatus according to a third embodiment. It is explanatory drawing for demonstrating an example of the utilization form of the information processing apparatus 30 which concerns on the modification 3. FIG. FIG. 10 is an explanatory diagram for describing an example of an information processing apparatus according to a fourth embodiment. FIG. 10 is an explanatory diagram for explaining another example of the information processing apparatus according to the fourth embodiment. FIG. 10 is an explanatory diagram illustrating an example of an information processing apparatus according to a fifth embodiment. 10 is an explanatory diagram for explaining an example of a schematic configuration in the vicinity of a lens of an imaging unit in an information processing apparatus according to Embodiment 5. FIG. It is a block diagram showing an example of functional composition of an information processor concerning a 2nd embodiment of this indication. It is explanatory drawing for demonstrating the basic principle of a process of a non-correlation component power estimation part. It is the figure which showed an example of the hardware constitutions of the signal processing apparatus which concerns on the same embodiment.

Hereinafter, preferred embodiments of the present disclosure will be described in detail with reference to the accompanying drawings. In addition, in this specification and drawing, about the component which has the substantially same function structure, duplication description is abbreviate | omitted by attaching | subjecting the same code | symbol.

The description will be made in the following order.
1. 1. First embodiment 1.1. Outline 1.2. Examination of installation location of sound collection section 1.3. Functional configuration 1.4. Processing 1.5. Example 1.5.1. Example 1: An example of a wearable device worn on the neck 1.5.2. Example 2: An example of a wearable device worn on the head 1.5.3. Example 3: Application example to portable information terminal 1.5.4. Example 4: Application example to a watch-type wearable device 1.5.5. Example 5: Application example to imaging apparatus Second Embodiment 2.1. Outline 2.2. Functional configuration 2.3. Details of Uncorrelated Component Power Estimator 2.4. Details of random noise power estimation section 2.5. Evaluation Hardware configuration Conclusion

<< 1. First Embodiment >>
<1.1. Overview>
First, an example of a schematic configuration of the information processing apparatus according to the first embodiment of the present disclosure will be described with reference to FIG. 1, and then a technical problem of the information processing apparatus according to the present embodiment will be described. . FIG. 1 is an explanatory diagram for describing an example of a schematic configuration of the information processing apparatus according to the first embodiment of the present disclosure.

In the example shown in FIG. 1, the information processing apparatus 10 is configured as a so-called wearable device. More specifically, the information processing apparatus 10 has a ring shape that is partially opened (in other words, a headband shape or a U-shape), and at least one of the inner surfaces of the ring-shaped portion. The part is attached to the user so as to contact a part of the user's neck (that is, to hang around the neck).

Further, the information processing apparatus 10 includes a sound collection unit such as a so-called microphone, and collects sound emitted by the user from the sound collection unit as acoustic information. For example, in the example illustrated in FIG. 1, the information processing apparatus 10 includes a plurality of sound collection units denoted by reference numerals 111 to 113. More specifically, the sound collection units 111 to 113 are supported by the housing 101 of the information processing apparatus 10, for example.

For example, FIG. 2 is an explanatory diagram for explaining an example of a schematic configuration of the information processing apparatus 10 according to the present embodiment, and a configuration of a portion of the information processing apparatus 10 in which the sound collection unit 111 is provided. It is the figure which showed an example. As shown in FIGS. 1 and 2, the information processing apparatus 10 has a streamlined shape so as to protrude toward the front of the user in the vicinity of the user's mouth when worn on the user's neck. The sound collecting portion 111 is provided at the tip of the convex portion (or in the vicinity of the tip) so as to face the direction in which the convex portion protrudes. Further, the sound collection unit 111 is configured as a device separate from the information processing apparatus 10, and faces the direction in which the convex portion protrudes with respect to the tip of the convex portion (or the vicinity of the tip). It may be supported. In the following description, when it is described that the sound collection unit 110 is provided for the information processing device 10, the sound collection unit 110 is separated from the information processing device 10 and the information processing device 10 is provided. The case where it is supported by at least a part of the device 10 can also be included.

Further, as shown in FIG. 1, the

sound collection units

112 and 113 are provided to face the information processing apparatus 10 in different directions. More specifically, the

sound collection units

112 and 113 are provided at positions that are substantially symmetric with respect to the user's neck when the information processing apparatus 10 is attached to the user's neck. Details of the position where each sound collecting unit is provided will be described later. In the example shown in FIG. 1, the

sound collecting portions

112 and 113 are provided so as to face the outside of the ring (that is, the side opposite to the center of the ring) with respect to the ring-shaped casing 101. ing. That is, the sound collection unit 112 and the sound collection unit 113 are provided so as to face opposite directions.

Based on such a configuration, for example, the information processing apparatus 10 uses the voice recognition technology or the natural language processing technology for the user's voice (acoustic information) collected by the sound collection unit (for example, the sound collection units 111 to 113). By performing an analysis based on the above, the content spoken by the user may be recognized. Thereby, the information processing apparatus 10 can recognize, for example, the instruction content from the user and execute various processes (applications) according to the recognition result.

As another example, the information processing apparatus 10 may have a so-called call function. In this case, the information processing apparatus 10 may transfer the sound collected by the sound collection unit (for example, the sound collection units 111 to 113) to another information processing apparatus that is a call partner.

On the other hand, the information processing apparatus 10 configured to be carried by the user like a so-called wearable device as shown in FIG. 1 has various usage scenes such as when used outdoors, and the information A situation is assumed in which the environment around the processing apparatus 10 changes dynamically. Under such circumstances, for example, randomly generated noise such as wind noise, noise due to vibration, and rubbing due to wearing of the apparatus is collected by the sound collecting unit of the information processing apparatus 10. There is a case.

Therefore, in the present disclosure, as an example of a mechanism that can collect the target sound in a more preferable manner even in an environment where noise is randomly generated, the installation position of each sound collection unit and the collection An example of signal processing based on the sound collection result by the sound unit will be described in detail.

<1.2. Examination of the location of the sound collector>
First, as shown in FIG. 1, the information processing apparatus 10 according to this embodiment is configured as a wearable device that is worn on the user's neck. The result of the examination regarding the installation position of the sound collecting unit that can be performed will be described. More specifically, it is assumed that a so-called wind noise is assumed to be noise, and each of the sound collection units when the wind is applied from different angles to the information processing apparatus 10 in which the sound collection units are installed at a plurality of locations. An example of the observation result of wind noise by will be described.

For example, FIG. 3 is an explanatory diagram for explaining an example of an observation environment for observing the effect of wind noise. In this observation, as shown in FIG. 3, the information processing apparatus 10 is mounted on the neck of a dummy doll U1 that imitates a part above the user's chest, and the circulator U2 is disposed in front of the dummy doll U1. Then, with the vertical direction of the dummy doll U1 as an axis, the dummy doll U1 is rotated in increments of 10 degrees within a range of 0 degrees to 360 degrees, so that the wind from the circulator U2 arrives at the information processing apparatus 10 The level of wind noise collected by each sound collecting unit was observed.

FIG. 4 shows an example of the installation positions of the plurality of sound collection units provided in the information processing apparatus 10 in this observation. Specifically, in the example shown in FIG. 4, sound collection units M1 to M6 are installed for the information processing apparatus 10. The markers attached to the information processing apparatus 10 schematically indicate the positions where the sound collecting units M1 to M6 are installed. In addition, in the marker to which the arrow was attached | subjected, the said arrow has shown the direction of the sound collection part corresponding to the said marker. For a marker without an arrow, the sound collection unit corresponding to the marker (that is, the sound collection units M3 and M6) is located in the vertical direction of the information processing apparatus 10 (that is, in the depth direction with respect to the drawing). It is set to face the front side.

Specifically, the sound collection unit M1 corresponds to the sound collection unit 111 in the information processing apparatus 10 described with reference to FIG. That is, the sound collection unit M1 is a convex portion provided so as to protrude toward the front of the user at a position corresponding to the vicinity of the user's mouth when the information processing apparatus 10 is worn by the user. It is provided at the tip. The sound collection unit M5 corresponds to the sound collection unit 112 in the information processing apparatus 10 described with reference to FIG. That is, when the information processing apparatus 10 is attached to the user, the sound collecting unit M5 is located at a position corresponding to the left side of the user (direction of approximately 270 degrees in FIG. 3). Outside the housing 101 (in other words, approximately 270 degrees in FIG. 3).

Further, the sound collection units M2 to M4 and M6 are positions corresponding to the area on the right front side of the user (in other words, the direction of approximately 45 degrees in FIG. 3) when the information processing apparatus 10 is worn by the user. Is provided. At this time, the sound collection unit M <b> 2 is interposed between the housing 101 of the information processing apparatus 10 and the user's neck and is installed so as to face the inside of the housing 101. The sound collection unit M4 is provided outside the housing 101 of the information processing apparatus 10 so as to face the outside of the housing 101 (in other words, the direction of approximately 45 degrees in FIG. 3). Note that the sound collection units M3 and M6 are provided so as to face vertically upward as described above.

FIG. 5 is an explanatory diagram for explaining an example of observation results of wind noise by each of the sound collection units when wind is applied to the information processing apparatus 10 from different angles. That is, FIG. 5 shows an example of a wind noise collection result by each of the sound collection units M1 to M6 described with reference to FIG. 4 in the observation environment described with reference to FIG. In the graph showing the sound collection results of the sound collection units M1 to M6 shown in FIG. 5, the numerical values described in the circumferential direction indicate the direction in which the wind from the circulator U2 arrives. The numerical values described along the radial direction of the graph indicate the level of the sound collected by the corresponding sound collection unit (that is, the observation level of the sound collection unit). That is, in the graph showing the sound collection results of the sound collection units M1 to M6 shown in FIG. 5, the smaller the observation level (in other words, the more the observation value is located inside the graph), the wind noise (ie, noise). ) Is less affected.

Here, paying attention to the observation result of the sound collection unit M1, it can be seen that the influence of the wind noise is small particularly in the situation where the wind comes from the front of the user (ie, the direction of 0 degree). In addition, regarding the sound collecting unit M1, it can be seen that the effect of wind noise is small compared to other sound collecting units even when the wind comes from directions other than the front.

For this reason, for example, as in the sound collection unit 111 shown in FIG. 1, the sound collection unit is provided at the tip of the streamlined projection (or in the vicinity of the tip) so that the projection protrudes. Thus, it is presumed that the influence of randomly generated noise such as wind noise can be reduced.

Further, when attention is paid to the observation results of the sound collection units M5 and M6, it is understood that the influence of wind noise is small when wind comes from the neck side of the user to the sound collection unit. This is presumed that the influence of wind noise is reduced by the wind being shielded by the user's neck and head.

From this, for example, like the

sound collection units

112 and 113 shown in FIG. 1, the user's part (for example, the neck or the head) to which the information processing apparatus 10 is attached can be used as a shield against wind or the like. By providing the sound collection unit, it is estimated that the characteristics of other sound collection units (for example, the sound collection unit 111 shown in FIG. 1) can be supplemented.

As described above, with reference to FIGS. 3 to 5, the information processing apparatus 10 according to the present embodiment is configured as a wearable device that is worn on the user's neck as an example. The result of the study on the installation position of the sound collection unit that can collect sound (ie, reduce the influence of noise such as wind noise) has been described.

<1.3. Functional configuration>
Next, with reference to FIG. 6, with respect to an example of the functional configuration of the information processing apparatus 10 according to the present embodiment, in particular, the information processing apparatus 10 uses a target sound (for example, A description will be given focusing on the process of acquiring the user's voice. FIG. 6 is a block diagram illustrating an example of a functional configuration of the information processing apparatus 10 according to the present embodiment.

As shown in FIG. 6, the information processing apparatus 10 includes a plurality of sound collecting units 111 to 11M (M is a positive integer), a frequency resolving unit 13, a channel power estimating unit 15, a filter estimating unit 16, a filter A processing unit 17 and a frequency synthesis unit 18 are included. In the following description, the sound collection units 111 to 11M may be referred to as “sound collection unit 110” unless they are particularly distinguished. Further, the number of sound collecting units 110 (that is, M) is not particularly limited as long as it is plural, but is more preferably 3 or more.

The sound collection unit 110 is configured as a sound collection device for collecting sound of the external environment (that is, sound that propagates through the external environment) like a so-called microphone. Note that the voice input from the user is also taken into the information processing apparatus 10 by being collected by the sound collection unit 110. The sound collection unit 110 may include a plurality of sound collection devices such as a microphone array. The sound collection unit 110 outputs an acoustic signal based on the sound collection result of the sound of the external environment to the frequency decomposition unit 13. Note that the acoustic signal output from the sound collection unit 110 may be input to the frequency decomposition unit 13 after the gain is adjusted by an amplifier or the like, converted from an analog signal to a digital signal by AD conversion, and the like. In the following description, when the channel number of the sound collection unit 110 is m (1 ≦ m ≦ M) and the discrete time is n, the acoustic signal output from the sound collection unit 110 is represented by x _m (n). It shall be expressed as

The frequency decomposing unit 13 is configured to decompose the acoustic signal x _m (n) output from the sound collecting unit 110 into frequency components and output the frequency components. Specifically, the frequency resolving unit 13 performs frame division, application of a predetermined window function, and time-frequency conversion (for example, FFT (Fast Fourier Transform), etc.) on the acquired acoustic signal x _m (n). The acoustic signal x _m (n) is decomposed into frequency components by performing processing such as DFT (Discrete Fourier Transform). In the following description, the frequency component of the acoustic signal x _m (n) may be described as X _m (i, k). Here, i indicates a frame number, and k indicates a discrete frequency number. Then, the frequency resolving unit 13 outputs each frequency component X _m (i, k) of the acquired acoustic signal x _m (n) to each of the filter processing unit 17 and the channel power estimating unit 15 located in the subsequent stage. To do. Thereby, for each of the sound collecting units 111 to 11M, each frequency component X _m (i, k) of the acoustic signal x _m (n) is output to the filter processing unit 17 and the channel power estimation unit 15, respectively. It will be.

The channel power estimation unit 15 acquires each frequency component X _m (i, k) of the acoustic signal x _m (n) for each sound collection unit 110 (that is, for each of the sound collection units 111 to 11M) from the frequency decomposition unit 13. To do. Next, the channel power estimation unit 15 determines the power of each sound collecting unit 110 for each frequency based on each frequency component X _m (i, k) of the acoustic signal x _m (n) corresponding to each sound collecting unit 110. Estimate the spectrum. Here, when the power spectrum corresponding to the i frame and the frequency k in the m-th sound collecting unit 110 (that is, the sound collecting unit 11m) is P _m (i, k), the power spectrum P _m (i, k) is expressed by a calculation formula shown as (Formula 1) below. In the following (Expression 1), X _m ^* (i, k) represents a conjugate complex number of X _m (i, k). In (Equation 1), r represents a smoothing coefficient in the frame direction for suppressing a rapid change in power spectrum (0 ≦ r <1).

... (Formula 1)

Then, the channel power estimation unit 15 outputs the estimation result of the power spectrum P _m (i, k) of each sound collection unit 110 to the filter estimation unit 16 for each frequency.

Based on the estimation result of the power spectrum P _m (i, k) of each sound collection unit 110 for each frequency output from the channel power estimation unit 15, the filter estimation unit 16 performs filtering after that. A filter coefficient for executing the process is calculated.

Specifically, the filter estimation unit 16 performs the following (Equation 2) based on the estimation result of the power spectrum P _m (i, k) of each sound collection unit 110 acquired from the channel power estimation unit 15 for each frequency. A matrix R (i, k) shown as follows is generated.

... (Formula 2)

Further, the filter estimation unit 16 determines, for each of the sound collection units 110, the frequency of each sound collection unit 110 based on the distance between the sound collection unit 110 and the target sound source (for example, the user's mouth). An array manifold vector a (k) showing attenuation and delay characteristics up to 110 is calculated. Note that the distance between the sound source of the target sound and each sound collection unit 110 is determined when the information processing apparatus 10 is attached to the user by the information processing apparatus 10 (and thus each information processing apparatus 10 provided with the information processing apparatus 10). It is possible to specify in advance based on the relative positional relationship between the sound collection unit 110) and the sound source.

Here, the array manifold vector a (k) is expressed by the following calculation formulas (Equation 3) and (Equation 4). Incidentally, in the calculation formula shown below, d _m is the target sound source (e.g., mouth) and, m-th sound collecting portion 110 (i.e., the sound collecting portion 11m) represents the distance between the. Further, g _m indicates the amount of attenuation until the target sound reaches the sound collecting unit 11m. Further, ω _k indicates an angular frequency corresponding to the discrete frequency number k. C indicates the speed of sound. A matrix with a superscript T is used to indicate transposition of the matrix. In the following description, a matrix with a superscript T may be referred to as a “transposed vector matrix”.

Based on the generated matrix R (i, k), the calculated array manifold vector a (k), and the condition shown below as (Equation 5), the filter estimation unit 16 The filter coefficient w (i, k) for executing the filtering process is calculated. Here, a matrix with a superscript H indicates a complex conjugate transpose of the matrix. In the following description, a matrix with a superscript H may be referred to as a “complex conjugate transposed vector matrix”.

... (Formula 5)

The filter coefficient w (i, k) for each frequency is expressed by the following calculation formula (Formula 6). Note that i indicates a frame number, and k indicates a discrete frequency number.

... (Formula 6)

Note that the filter coefficient w (i, k) shown as (Equation 6) above indicates the gain of the component a (k) coming from the target sound source (for example, the mouth), as shown in (Equation 5) above. 1 and a coefficient that minimizes noise components (for example, wind noise). Then, the filter estimation unit 16 outputs the filter coefficient w (i, k) calculated for each frequency to the filter processing unit 17.

The filter processing unit 17 acquires each frequency component X _m (i, k) of the acoustic signal x _m (n) for each sound collection unit 110 (ie, for each of the sound collection units 111 to 11M) from the frequency decomposition unit 13. . Further, the filter processing unit 17 acquires the filter coefficient w (i, k) calculated for each frequency from the filter estimation unit 16. The filter processing unit 17 uses each frequency component X _m (i, k) of the acoustic signal x _m (n) for each sound collecting unit 110 as an input signal and is based on the acquired filter coefficient w (i, k) for each frequency. By performing the filtering process, an output signal Y (i, k) is generated.

Specifically, the filter processing unit 17 uses each frequency component X _m (i, k) of the acoustic signal x _m (n) for each sound collecting unit 110 as an input signal, and obtains a filter coefficient w (i for each acquired frequency. , K), the input signal is weighted and added to generate an output signal Y (i, k) for each frequency. For example, the output signal Y (i, k) is expressed by a calculation formula shown as (Formula 7) below. Note that i indicates a frame number, and k indicates a discrete frequency number.

... (Formula 7)

Then, the filter processing unit 17 outputs the output signal Y (i, k) generated for each frequency to the frequency synthesis unit 18.

The frequency synthesizer 18 acquires the output signal Y (i, k) generated for each frequency from the filter processor 17. The frequency synthesizer 18 generates the acoustic signal y (n) by synthesizing the acquired output signal Y (i, k) for each frequency. That is, the frequency synthesizing unit 18 performs the reverse process of the frequency decomposing unit 13 described above. Specifically, the frequency synthesizer 18 performs frequency-time conversion (for example, IFFT (Inverse FFT), IDFT (Inverse DFT), etc.), a predetermined window for the output signal Y (i, k) for each frequency. By performing processing such as function application and frame synthesis, an acoustic signal y (n) in which the output signal Y (i, k) for each frequency is synthesized is generated.

For example, FIG. 7 shows an example of processing in which the information processing apparatus 10 according to the present embodiment acquires the target sound based on the sound collection results of the plurality of sound collection units 110. In the example illustrated in FIG. 7, an example in which four microphones of the sound collection units 111 to 114 are used as the plurality of sound collection units 110 is illustrated. That is, in the example shown in FIG. 7, each of the sound collection units 111 to 114 is obtained by an example of the sound collection result by each of the sound collection units 111 to 114 (ie, the collected sound signal) and the signal processing by the information processing apparatus 10. An example of an acoustic signal (synthetic sound) obtained by synthesizing the sound collection results is shown.

As described above, the filtering process for synthesizing the sound collection results (more specifically, each frequency component X _m (i, k) of the acoustic signal x _m (n)) by each of the plurality of sound collection units 110. The coefficient w (i, k) is a characteristic that keeps the gain of the component a (k) coming from the sound source (for example, the mouth) of the target sound at 1 and minimizes the noise component (for example, wind noise). Have With such a configuration, weighting is performed so that the input of the sound collecting unit 110 having a smaller noise component level (in other words, the sound collecting unit 110 having a smaller noise component effect) is prioritized, and each sound collecting unit 110 is prioritized. The sound collection results are synthesized. By such processing, even in an environment where noise such as wind noise is generated randomly, it is possible to suppress the influence of the noise and collect the target sound in a more preferable manner.

As described above, the information processing apparatus 10 according to the present embodiment is configured to synthesize the target sound from the sound collection results of each of the plurality of sound collection units 110. This is different from the configuration in which the sound collection unit 110 that obtains the sound collection result is simply switched. More specifically, in the case of a configuration in which the sound collecting unit 110 that acquires the sound collection result is simply switched, the acoustic signal may be deteriorated before and after the switching, and particularly noise such as wind noise. In a situation where the direction of arrival of the sound changes dynamically, the deterioration of the acoustic signal tends to be more obvious. On the other hand, the information processing apparatus 10 according to the present embodiment synthesizes the target sound by the signal processing described above, so that the direction in which noise such as wind noise arrives dynamically changes. In addition, the target sound can be acquired in a more natural manner without causing deterioration of the acoustic signal.

The signal processing for the sound collection result of each sound collection unit 110 described above is merely an example, and weighting is performed so that the input of the sound collection unit 110 having a smaller noise component level is prioritized, and each sound collection unit 110 is prioritized. The content is not particularly limited as long as the sound collection results of the sound unit 110 can be synthesized.

Then, the frequency synthesizer 18 outputs the generated acoustic signal y (n) as a target sound collection result. The acoustic signal y (n) output from the frequency synthesizer 18 is used for various processes (for example, voice recognition, voice call, etc.) executed by the information processing apparatus 10, for example.

Note that the configuration illustrated in FIG. 6 is merely an example, and the configuration of the information processing apparatus 10 is not necessarily limited to the example illustrated in FIG. 6 as long as the various processes described above can be realized. For example, in the example shown in FIG. 6, the frequency resolving unit 13 is provided for each of the sound collecting units 111 to 11m. It is good also as a structure to process. A part of the configuration may be externally attached to the information processing apparatus 10. As a specific example, at least some of the plurality of sound collecting units 110 may be configured to be detachable from the information processing apparatus 10.

As described above, with reference to FIG. 6 and FIG. 7, with regard to an example of the functional configuration of the information processing apparatus 10 according to the present embodiment, in particular, the information processing apparatus 10 has the target sound based on the sound collection results of the plurality of sound collection units. The description has been given focusing on the process of acquiring the.

<1.4. Processing>
Next, referring to FIG. 8, an example of a flow of a series of processes of the information processing apparatus 10 according to the present embodiment, in particular, the information processing apparatus 10 is based on the sound collection results of each of the plurality of sound collection units. Description will be made by paying attention to processing for acquiring sound (for example, user's voice). FIG. 8 is a flowchart illustrating an example of a flow of a series of processes of the information processing apparatus 10 according to the present embodiment.

(Step S101)
The sound of the external environment is taken into the information processing apparatus 10 by being collected by the plurality of sound collection units 110. The sound collection unit 110 adjusts the gain of an acoustic signal (analog signal) based on the sound collection result, converts the analog signal from a digital signal by AD conversion, and then converts the converted acoustic signal (digital signal) x _m (n ) Is output to the frequency resolving unit 13.

(Step S103)
The frequency resolving unit 13 performs processing such as frame division, application of a predetermined window function, time-frequency conversion, and the like on the acoustic signal x _m (n) output from the sound collecting unit 110. The acoustic signal x _m (n) is decomposed into frequency components. Then, the frequency resolving unit 13 outputs each frequency component X _m (i, k) of the acoustic signal x _m (n) to each of the filter processing unit 17 and the channel power estimating unit 15 located in the subsequent stage. Thereby, for each of the plurality of sound collecting units 110, each frequency component X _m (i, k) of the acoustic signal x _m (n) is output to the filter processing unit 17 and the channel power estimation unit 15, respectively. It will be.

(Step S105)
The channel power estimation unit 15 acquires each frequency component X _m (i, k) of the acoustic signal x _m (n) from the frequency resolution unit 13 for each sound collection unit 110. Next, the channel power estimation unit 15 determines the power of each sound collecting unit 110 for each frequency based on each frequency component X _m (i, k) of the acoustic signal x _m (n) corresponding to each sound collecting unit 110. Estimate the spectrum. Then, the channel power estimation unit 15 outputs the estimation result of the power spectrum P _m (i, k) of each sound collection unit 110 to the filter estimation unit 16 for each frequency.

(Step S107)
Based on the estimation result of the power spectrum P _m (i, k) of each sound collection unit 110 for each frequency output from the channel power estimation unit 15, the filter estimation unit 16 performs filtering after that. A filter coefficient w (i, k) for executing the processing is calculated.

Specifically, the filter estimation unit 16 generates a matrix R (i, k) based on the power spectrum P _m (i, k) of each sound collection unit 110. Further, the filter estimation unit 16 determines, for each frequency, the attenuation and delay characteristics to the sound collection unit 110 based on the distance between the sound collection unit 110 and the target sound source for each frequency. The indicated array manifold vector a (k) is calculated. Then, the filter estimation unit 16 calculates a filter coefficient w (i, k) based on the generated matrix R (i, k) and the calculated array manifold vector a (k), and the filter coefficient w (i , K) is output to the filter processing unit 17.

(Step S109)
The filter processing unit 17 acquires each frequency component X _m (i, k) of the acoustic signal x _m (n) for each sound collection unit 110 from the frequency decomposition unit 13. Further, the filter processing unit 17 acquires the filter coefficient w (i, k) calculated for each frequency from the filter estimation unit 16. The filter processing unit 17 uses each frequency component X _m (i, k) of the acoustic signal x _m (n) for each sound collecting unit 110 as an input signal, based on the acquired filter coefficient w (i, k) for each frequency. By weighting and adding the input signals, an output signal Y (i, k) for each frequency is generated. Then, the filter processing unit 17 outputs the output signal Y (i, k) generated for each frequency to the frequency synthesis unit 18.

(Step S111)
The frequency synthesizer 18 performs processing such as frequency-time conversion, application of a predetermined window function, and frame synthesis on the output signal Y (i, k) for each frequency output from the filter processor 17. Thus, the output signal Y (i, k) for each frequency is synthesized. As a result, an acoustic signal y (n) in which the sound collection results by the sound collection units 110 are combined is generated. The acoustic signal y (n) generated by the frequency synthesizer 18 is used as a sound collection result for various processes (for example, voice recognition, voice call, etc.) executed by the information processing apparatus 10. .

As described above, with reference to FIG. 8, with regard to an example of a series of processing flow of the information processing apparatus 10 according to the present embodiment, in particular, the information processing apparatus 10 performs the target sound based on the sound collection results of the plurality of sound collection units. The description has been given focusing on the process of acquiring the.

<1.5. Example>
Next, another example of the information processing apparatus 10 according to the present embodiment will be described as an example.

<1.5.1. Example 1: An example of a wearable device attached to the neck>
First, referring to FIG. 9 to FIG. 11, as an example 1, an information processing apparatus configured as a wearable device that can be worn on the user's neck like the so-called neckband type wearable device shown in FIG. An example will be described.

For example, FIG. 9 is an explanatory diagram for describing an example of the information processing apparatus according to the first embodiment, and illustrates an example of the information processing apparatus configured as a wearable device that can be worn on the user's neck. In this description, the information processing apparatus shown in FIG. 9 is referred to as an “information processing apparatus 10a” in order to distinguish it from the information processing apparatus 10 according to the above-described embodiment and the information processing apparatuses according to other examples. There is a case.

As shown in FIG. 9, the information processing apparatus 10a includes sound collection units 111 to 114. The sound collection units 111 to 113 correspond to the sound collection units 111 to 113 in the information processing apparatus 10 described above with reference to FIG. Further, the sound collecting unit 114 is provided at a position behind the user so that the information processing apparatus 10a is attached to the user's neck so as to face the rear side of the user. With such a configuration, for example, it is possible to further reduce the influence of noise coming from behind the user.

Further, the information processing apparatus 10a is provided with convex portions having a streamlined shape protruding in the direction in which each of the sound collecting portions 112 to 114 faces at a position where the sound collecting portions 112 to 114 are installed. Each of the sound collecting portions 112 to 114 is provided at the tip. By adopting such a configuration, the sound collecting portions 112 to 114 also reduce the influence of noise such as wind noise as in the sound collecting portion 111, and the direction in which the convex portion protrudes (that is, the sound collecting portion). It is possible to collect sound coming from the direction in which the sound part is directed in a more preferable manner.

It should be noted that the position where the convex portion is provided (that is, the position where the sound collecting portion 110 is provided) is not particularly limited. Therefore, for example, by providing various circuits such as a driver, a battery, and the like, a convex portion is provided at a location where the casing 101 may swell, and the sound collection unit 110 is provided at the tip of the convex portion (or in the vicinity of the tip). It is good also as a structure which provides.

FIG. 10 is an explanatory diagram for explaining another example of the information processing apparatus according to the first embodiment, and illustrates an example of the information processing apparatus configured as a wearable device that can be worn on the user's neck. . In this description, the information processing apparatus shown in FIG. 10 is referred to as an “information processing apparatus 10b” in order to distinguish it from the information processing apparatus 10 according to the above-described embodiment and the information processing apparatuses according to other examples. There is a case.

As shown in FIG. 10, the information processing apparatus 10b has a ring shape and is configured to be openable at a portion indicated by reference numeral 19. In addition, it is comprised so that attachment or detachment is possible between the edge parts which mutually spaces apart because the part shown with the referential mark 19 opens. With such a configuration, the information processing apparatus 10b is attached to the user so that the inner surface of the ring-shaped portion is in contact with the user's neck (that is, wound around the neck).

Further, the information processing apparatus 10b has the sound collecting units 115 to 118 at positions different from each other along the circumference of the ring-shaped casing (ie, on the side opposite to the center of the ring). It is provided to face. In the information processing apparatus 10b, the sound collection units 115 to 118 correspond to the sound collection unit 110 (for example, the sound collection units 111 to 113 shown in FIG. 1) in the information processing apparatus 10 according to the above-described embodiment. To do.

With this configuration, each of the sound collection units 115 to 118 shields noise coming from the opposite side to the direction in which the sound collection unit is facing by the user's part (that is, the neck) to which the information processing apparatus 10b is attached. Therefore, the influence of the noise is mitigated. In particular, in the information processing apparatus 10b shown in FIG. 10, each of the sound collection units 115 to 118 is supported so as to be closer to the user's neck than the information processing apparatus 10 shown in FIG. The influence of noise such as noise (particularly noise coming from the user's neck) is further alleviated. This is because the influence of noise coming from the user's part side is further alleviated in the sound collecting parts M5 and M6 (that is, the sound collecting part closer to the user's part) described with reference to FIG. It is clear from that. In addition, since each of the sound collection units 115 to 118 is provided so as to face different directions, for example, the characteristics of other sound collection units may be supplemented based on the sound collection results of some sound collection units. It becomes possible.

In the information processing apparatus 10b shown in FIG. 10 as well, a streamline-shaped convex portion is provided on at least a part of the housing, and a sound collecting unit 110 (for example, at the vicinity of the convex portion) , At least a part of the sound collecting units 115 to 118) may be provided.

FIG. 11 is an explanatory diagram for explaining another example of the information processing apparatus according to the first embodiment, and shows an example of the information processing apparatus configured as a wearable device having a shape like a so-called necklace. Yes. In this description, the information processing apparatus shown in FIG. 11 is referred to as an “information processing apparatus 10c” in order to distinguish it from the information processing apparatus 10 according to the above-described embodiment and the information processing apparatuses according to other examples. There is a case.

11, reference numeral 119 indicates an example of the sound collection unit 110 in the information processing apparatus 10 according to the above-described embodiment. That is, in the information processing apparatus 10c having a shape like a necklace, for example, a streamline-shaped convex portion is provided in a portion corresponding to a so-called pendant so as to face the front of the user when the user wears the convex portion. The sound collecting unit 119 may be provided at the tip of the head (or in the vicinity of the tip).

In the example shown in FIG. 11, one sound collection unit 110 is provided for the information processing apparatus 10c, but a plurality of sound collection units 110 may be provided. Further, when a plurality of sound collecting units 110 are provided for the information processing apparatus 10c, the plurality of sound collecting units 110 may be provided so as to face different directions.

As described in the first embodiment, an information processing apparatus configured as a wearable device that can be worn on the user's neck like the so-called neckband type wearable device shown in FIG. An example has been described.

<1.5.2. Example 2: Example of wearable device worn on head>
Next, as a second embodiment, an example of an information processing apparatus configured as a wearable device that can be worn on the head will be described with reference to FIGS.

For example, FIG. 12 is an explanatory diagram for explaining an example of the information processing apparatus according to the second embodiment, and illustrates an example of the information processing apparatus configured as a wearable device that can be worn on the user's head. In this description, the information processing apparatus shown in FIG. 12 is referred to as an “information processing apparatus 20a” in order to distinguish it from the information processing apparatus 10 according to the above-described embodiment and the information processing apparatuses according to other examples. There is a case.

As shown in FIG. 12, the information processing apparatus 20a is mounted on the user's head so that a housing in which circuits and the like for realizing various functions are incorporated is held near the user's ear. . As a specific example, in the example illustrated in FIG. 12, the information processing device 20 a includes an earphone unit that is inserted into the user's ear canal and a cable-shaped support member that supports the housing by being hooked on the user's ear. Prepare. In the information processing apparatus 20a, the housing is held near the user's ear by the earphone unit and the cable-shaped support member.

As shown in FIG. 12, the information processing apparatus 20a includes

sound collection units

211 and 212. In the information processing apparatus 20a, the

sound collection units

211 and 212 are connected to the sound collection unit 110 (for example, the sound collection units 111 to 113 shown in FIG. 1) in the information processing apparatus 10 according to the above-described embodiment. Equivalent to.

Specifically, the information processing apparatus 20a, when attached to the user's head, places the front side on an end portion of the casing that is held near the user's ear and located on the front side of the user. A convex portion having a streamline shape protruding so as to face is provided. And the sound-collecting part 211 is provided in the front-end | tip of the said convex part so that the direction (namely, user's front) where the said convex part protrudes. In addition, when the information processing device 20a is attached to the user's head, the information processing device 20a has at least a part of a side surface located outside the housing (that is, opposite to the head) in the direction of the outside ( That is, the sound collection unit 212 is provided so as to face the user's lateral direction. In addition, the information processing apparatus 20a includes a convex portion having a streamline shape that protrudes toward the outer side of the casing with respect to the side surface of the casing, and the sound collection unit 212 is provided at the tip of the convex portion. It may be provided.

In the example illustrated in FIG. 12, the description has been given focusing on the housing held near the user's left ear, but the housing held near the user's right ear is also located near the left ear. A configuration similar to the case to be held can be taken. Specifically, the housing held on the right ear side may be provided with only a configuration corresponding to the sound collection unit 212, or a configuration corresponding to the

sound collection units

211 and 212. Also good.

FIG. 13 is an explanatory diagram for explaining another example of the information processing apparatus according to the second embodiment. The information processing apparatus is configured as a so-called glasses-type wearable device that is worn on the user's head. An example is shown. In this description, the information processing apparatus shown in FIG. 13 is referred to as an “information processing apparatus 20b” in order to distinguish it from the information processing apparatus 10 according to the above-described embodiment and the information processing apparatuses according to other examples. There is a case.

As shown in FIG. 13, the information processing apparatus 20b includes sound collection units 213 to 215. In the information processing apparatus 20b, the sound collection units 213 to 215 are connected to the sound collection unit 110 (for example, the sound collection units 111 to 113 shown in FIG. 1) in the information processing apparatus 10 according to the embodiment described above. Equivalent to.

For example, in the information processing apparatus 20b, a sound collection unit 213 is provided at least at a part corresponding to the front of the glasses. As a more specific example, the information processing apparatus 20b includes a convex portion having a streamline shape that protrudes forward at a portion corresponding to a bridge of glasses, and a sound collecting portion is provided at the tip of the convex portion. 213 is provided so as to face the direction in which the convex portion protrudes. Further, as another example, as shown by reference numeral 213 ′, the convex portion and the sound collecting portion are provided in a portion corresponding to the front of the glasses other than the portion corresponding to the bridge. May be.

Further, the information processing apparatus 20b is provided with

sound collection units

214 and 215 at least at a part corresponding to the temples of the glasses. The

sound collection units

214 and 215 are provided so as to face in the direction opposite to the head (that is, the lateral direction of the user) when the information processing apparatus 20b is mounted on the user's head, for example. It is good to have been.

FIG. 14 is an explanatory diagram for explaining another example of the information processing apparatus according to the second embodiment. Another example of the information processing apparatus configured as a wearable device that can be worn on the user's head. Show. In this description, the information processing apparatus shown in FIG. 14 is referred to as an “information processing apparatus 20c” in order to distinguish it from the information processing apparatus 10 according to the above-described embodiment and the information processing apparatuses according to other examples. There is a case.

As shown in FIG. 14, the information processing apparatus 20c includes sound collection units 216 to 218. In the information processing apparatus 20c, the sound collection units 216 to 218 are connected to the sound collection unit 110 (for example, the sound collection units 111 to 113 shown in FIG. 1) in the information processing apparatus 10 according to the embodiment described above. Equivalent to.

More specifically, the sound collecting portions 216 to 218 are provided at different positions of the portions corresponding to the frame of the glasses (for example, the front and the temple) so as to face different directions. More specifically, each of the sound collection units 216 to 218 is provided so as to face the direction opposite to the head when the information processing apparatus 20c is attached to the user's head.

With such a configuration, each of the sound collecting units 216 to 218 is shielded by the user's head from the noise coming from the side opposite to the direction in which the sound collecting unit 216 is facing, so that the influence of the noise is not affected. Alleviated. In addition, since each of the sound collection units 216 to 218 is provided so as to face different directions, for example, the characteristics of other sound collection units may be supplemented based on the sound collection results of some sound collection units. It becomes possible.

FIG. 15 is an explanatory diagram for explaining another example of the information processing apparatus according to the second embodiment, and illustrates an example of the information processing apparatus configured as an overhead wearable device such as a so-called headphone. ing. In this description, the information processing apparatus shown in FIG. 15 is referred to as an “information processing apparatus 20d” in order to distinguish it from the information processing apparatus 10 according to the above-described embodiment and the information processing apparatuses according to other examples. There is a case.

In the example illustrated in FIG. 15, the information processing apparatus 20 d includes an imaging unit 25 and a sound collection unit 219. In the information processing apparatus 20d, the sound collection unit 219 corresponds to the sound collection unit 110 (for example, the sound collection units 111 to 113 shown in FIG. 1) in the information processing apparatus 10 according to the above-described embodiment. .

Specifically, when the information processing device 20d is mounted on the user's head, the imaging unit 25 can store the front of the user of the housing of the information processing device 20d within the angle of view. In the position. For example, in the example illustrated in FIG. 15, the imaging unit 25 is provided in the housing of the information processing device 20 d so as to face the front of the user.

Further, the information processing apparatus 20d includes a convex portion having a streamline shape that protrudes toward the front side of the user in a state of being mounted on the user's head in at least a part of the housing, and the convex portion A sound collecting portion 219 is provided at the front end of the head so as to face the direction in which the convex portion protrudes. For example, in the example illustrated in FIG. 15, the sound collection unit 219 is provided in the vicinity of the imaging unit 25. Further, as another example, as indicated by reference numeral 219 ′, a streamline projecting so as to face the front side of the user on at least a part of the holding member for holding the information processing apparatus 20d on the head of the user. And a sound collecting part may be provided at the tip of the convex part so as to face the direction in which the convex part protrudes.

As described above, as the second embodiment, an example of an information processing apparatus configured as a wearable device that can be worn on the head has been described with reference to FIGS. 12 to 15. In addition, the example demonstrated above is an example to the last, and is not necessarily limited to the example shown above. As a specific example, a configuration corresponding to the sound collection unit 110 in the information processing apparatus 10 according to the above-described embodiment with respect to the information processing apparatus configured as a head-mounted wearable device having a so-called headband shape. May be provided.

<1.5.3. Example 3: Application example to portable information terminal>
Next, as Example 3, an example of an information processing apparatus configured as a portable information terminal such as a so-called smartphone will be described with reference to FIGS. 16 and 17.

For example, FIG. 16 is an explanatory diagram for explaining an example of the information processing apparatus according to the third embodiment. In this description, the information processing apparatus shown in FIG. 16 is referred to as an “information processing apparatus 30” in order to distinguish it from the information processing apparatus 10 according to the above-described embodiment and the information processing apparatuses according to other examples. There is a case.

As shown in FIG. 16, the information processing apparatus 30 includes sound collection units 311 to 314. In the information processing apparatus 30, the sound collection units 311 to 314 are connected to the sound collection unit 110 (for example, the sound collection units 111 to 113 shown in FIG. 1) in the information processing apparatus 10 according to the above-described embodiment. Equivalent to.

Specifically, the housing of the information processing apparatus 30 has a substantially rectangular surface 36 at least in part, and a predetermined region including the corner of the surface 36 (that is, the corner or the vicinity of the corner). ), A convex portion having a streamline shape is formed so as to face the outside of the housing. In other words, the housing of the information processing apparatus 30 includes a substantially planar surface 36 and a plurality of side surfaces 371 to 374 formed so as to face different directions along the end of the surface 36, and are adjacent to each other. A convex portion having a streamline shape is formed in a predetermined region including a portion where the matching side surfaces are connected. The surface 36 may correspond to a surface on which a display unit such as a display is provided, for example. Further, the corner itself of the housing of the information processing device 30 may be the convex portion. Each of the sound collection units 311 to 314 is provided at the tip of any one of the convex portions (or in the vicinity of the tip) so as to face the outside of the casing of the information processing apparatus 30.

FIG. 17 is an explanatory diagram for explaining an example of a usage form of the information processing apparatus 30 according to the modification 3. An example of a case where the user performs a voice call using the information processing apparatus 30 is illustrated. Show.

As shown in FIG. 17, for example, when the user makes a voice call while holding the information processing apparatus 30 in the vicinity of his / her right ear, the information is set so that the sound collecting unit 312 faces the front of the user. The processing device 30 is held. With such a configuration, for example, even in a situation where the user makes a voice call while moving, the sound collecting unit 312 is affected by wind noise caused by the wind coming from the front due to the movement of the user. It becomes difficult. It can be assumed that the user makes a voice call while holding the information processing apparatus 30 in the vicinity of his / her left ear. In this case, the information processing apparatus 30 is held so that the sound collecting unit 311 faces substantially in front of the user, and the sound collecting unit 311 winds off due to wind coming from the front due to the movement of the user. Less affected by sound. That is, the information processing apparatus 30 can mitigate the influence of wind noise caused by the wind coming from the front due to the movement of the user based on the configuration described above.

In addition, the information processing apparatus 30 is provided so that the sound collecting units 311 to 314 face different directions. With such a configuration, the information processing apparatus 30 can supplement the characteristics of the other sound collection units based on the sound collection results of at least some of the sound collection units.

As described above, as an example 3, an example of an information processing apparatus configured as a portable information terminal such as a so-called smartphone has been described with reference to FIGS. 16 and 17.

<1.5.4. Example 4: Application example to watch-type wearable device>
Next, as Example 4, an example of an information processing apparatus configured as a so-called watch-type wearable device that can be worn on an arm will be described with reference to FIGS. 18 and 19.

For example, FIG. 18 is an explanatory diagram for explaining an example of the information processing apparatus according to the fourth embodiment. In this description, the information processing apparatus shown in FIG. 18 is referred to as an “information processing apparatus 40a” in order to distinguish it from the information processing apparatus 10 according to the above-described embodiment and the information processing apparatuses according to other examples. There is a case.

As shown in FIG. 18, the information processing apparatus 40a includes sound collection units 411 to 415. In the information processing apparatus 30, the sound collection units 411 to 415 are connected to the sound collection unit 110 (for example, the sound collection units 111 to 113 shown in FIG. 1) in the information processing apparatus 10 according to the above-described embodiment. Equivalent to.

Specifically, the information processing apparatus 40a includes a casing 481 in which circuits for realizing various functions are incorporated, and a belt-like support member 482 that supports the casing 481 on the user's arm. Similar to the information processing apparatus 30 according to the third embodiment described above, the casing 481 has at least a part of a substantially rectangular surface, and the casing 481 has a predetermined area including corners of the substantially rectangular surface. A convex portion having a streamline shape is formed so as to face the outside of the body 481. Note that the substantially rectangular surface corresponds to a surface on the side where a dial in a so-called timepiece is provided. Each of the sound collecting portions 411 to 414 is provided at the tip of one of the convex portions (or in the vicinity of the tip) so as to face the outside of the housing 481.

In addition, in the state where the information processing apparatus 40a is attached to the arm, the support member 482 includes a sound collection unit 415 at a position that is substantially symmetric with respect to the housing 481 with respect to the arm, and is opposite to the arm. It is provided so as to face the direction.

With such a configuration, the information processing apparatus 40a allows at least any one of the sound collection units 411 to 414 even under a situation where the user is waving the arm on the side on which the information processing apparatus 40a is mounted. However, it is in a state of facing the direction substantially equal to the direction in which the arm is swung. Therefore, the information processing apparatus 40a can mitigate the effect of wind noise caused by arm swing based on the sound collection results of the sound collection units 411 to 414. In addition, the information processing apparatus 40a is provided so that the sound collecting units 411 to 415 face different directions. In particular, with respect to the sound collection unit 415, noise arriving from the opposite side to the direction in which the sound collection unit 415 faces is shielded by the arm on which the information processing device 40a is attached. With such a configuration, the information processing apparatus 40a can supplement the characteristics of other sound collection units based on the sound collection results of at least some of the sound collection units 411 to 415.

FIG. 19 is an explanatory diagram for explaining another example of the information processing apparatus according to the fourth embodiment. In this description, the information processing apparatus shown in FIG. 19 is referred to as an “information processing apparatus 40b” in order to distinguish it from the information processing apparatus 10 according to the above-described embodiment and the information processing apparatuses according to other examples. There is a case.

As shown in FIG. 19, the information processing apparatus 40b includes a sound collecting unit 416 in a portion corresponding to a so-called screw portion of a timepiece indicated by reference numeral 483 (hereinafter referred to as “screw portion 483”). Specifically, the screw part 483 may be used as a convex part for providing the sound collection part 416 by forming the screw part 483 to have a streamlined shape. In the information processing apparatus 40b, the sound collection unit 416 corresponds to the sound collection unit 110 (for example, the sound collection unit 111) in the information processing apparatus 10 according to the above-described embodiment.

As described above, as the fourth embodiment, an example of the information processing apparatus configured as a so-called watch-type wearable device that can be worn on the arm has been described with reference to FIGS. 18 and 19.

<1.5.5. Example 5: Application Example to Imaging Device>
Next, as Example 5, an example of an information processing apparatus configured as an imaging apparatus capable of capturing a moving image or a still image will be described with reference to FIGS. 20 and 21. FIG.

For example, FIG. 20 is an explanatory diagram for explaining an example of the information processing apparatus according to the fifth embodiment. In this description, the information processing apparatus shown in FIG. 20 is referred to as an “information processing apparatus 50” in order to distinguish it from the information processing apparatus 10 according to the above-described embodiment and the information processing apparatuses according to other examples. There is a case.

In FIG. 20, reference numeral 53 corresponds to an image capturing unit for capturing images such as moving images and still images.

Reference numerals

511 and 512 correspond to an example of a sound collection unit provided in the information processing apparatus 50. In the information processing apparatus 50, the

sound collection units

511 and 512 are connected to the sound collection unit 110 (for example, the sound collection units 111 to 113 shown in FIG. 1) in the information processing apparatus 10 according to the embodiment described above. Equivalent to.

Specifically, as illustrated in FIG. 20, the information processing device 50 includes, for example, a direction of the casing that supports the imaging unit 53 in which the imaging unit 53 captures an image (hereinafter referred to as “imaging direction”). A convex portion having a streamline shape protruding in the imaging direction is provided on a part of the surface facing (which may be). A sound collection unit 511 is provided at the tip of the convex portion (or in the vicinity of the tip) so as to face the imaging direction of the imaging unit 53 (in other words, forward).

Further, a sound collection unit 512 may be provided in the vicinity of the imaging unit 53 (for example, in the vicinity of the lens of the imaging unit 53). For example, FIG. 21 is an explanatory diagram for explaining an example of a schematic configuration in the vicinity of the lens of the imaging unit 53 in the information processing apparatus 50 according to the fifth embodiment. In the example illustrated in FIG. 21, the information processing device 50 is provided with a convex portion 551 that protrudes toward the outside of the housing of the information processing device 50 in the vicinity of the lens of the imaging unit 53. Further, the convex portion 551 includes a convex portion 553 having a streamline shape that protrudes in the imaging direction (that is, the front side) of the imaging unit 53, and is collected at the tip of the convex portion 553 (or in the vicinity of the tip). A sound unit 513 is provided.

With such a configuration, the information processing apparatus 50 can mitigate the effect of wind noise caused by the wind coming from the front due to the movement of the user, for example, even in a situation where an image is captured while the user is moving. Is possible.

Although not shown in FIGS. 20 and 21, the information processing apparatus 50 may include another sound collection unit different from the

sound collection units

511 and 512. In this case, the other sound collecting unit may be provided so as to face a different direction from the

sound collecting units

511 and 512. As a more specific example, for example, the housing of the information processing device 50 is directed to the surface opposite to the imaging direction of the imaging unit 53 and to the direction opposite to the imaging direction (that is, rearward). The other sound collecting unit may be provided. With such a configuration, for example, the characteristics of the

sound collection units

511 and 512 can be supplemented based on the sound collection result by another sound collection unit.

As described above, as Example 5, an example of an information processing apparatus configured as an imaging apparatus capable of capturing a moving image or a still image has been described with reference to FIGS. 20 and 21.

<< 2. Second Embodiment >>
<2.1. Overview>
Subsequently, a second embodiment of the present disclosure will be described. In the information processing apparatus 10 according to the first embodiment described above, the input of the sound collection unit having a smaller observation level (that is, the level of the collected sound) is received based on the sound collection results of each of the plurality of sound collection units. By performing the filtering process so as to be prioritized, the influence of noise generated randomly such as wind noise has been reduced. Such control makes it possible to mitigate the influence of the noise in a more preferable manner, particularly when the influence of randomly generated noise such as wind noise is larger.

On the other hand, when the sound collection result of each sound collection unit is evaluated as it is as in the above-described control, the target sound is more accurately obtained in a situation where the target sound such as speech is collected as a main component. The sound collection result of the sound collection unit collected at a high level may not be used. That is, in the situation where the influence of randomly generated noise such as wind noise is small, for example, the sound collection result of the sound collection unit having a small signal-to-noise ratio is preferentially used. There is.

Therefore, the present embodiment maintains the effect of suppressing noise generated randomly such as wind noise as in the first embodiment described above, and is more preferable when the influence of randomly generated noise is small. An example of a mechanism that can acquire the target sound in various ways is proposed.

<2.2. Functional configuration>
First, an example of a functional configuration of the information processing apparatus according to the present embodiment will be described with reference to FIG. FIG. 22 is a block diagram illustrating an example of a functional configuration of the information processing apparatus according to the present embodiment. In the following description, in order to explicitly distinguish the information processing apparatus according to the present embodiment from the information processing apparatus 10 according to the first embodiment described above (see FIG. 6), “information processing apparatus 60”. May be called.

As shown in FIG. 22, the information processing apparatus 60 according to the present embodiment includes a plurality of sound collecting units 111 to 11M (M is a positive integer), a frequency resolving unit 13, a channel power estimating unit 65, and filter estimation. A unit 66, a filter processing unit 17, and a frequency synthesis unit 18 are included. The plurality of sound collecting units 111 to 11M (M is a positive integer), the frequency resolving unit 13, the filter processing unit 17, and the frequency synthesizing unit 18 are the information processing apparatus according to the first embodiment described above. 10 (refer to FIG. 6) corresponds to the configuration with the same reference numerals. That is, the information processing apparatus 60 according to the present embodiment is different from the information processing apparatus 10 according to the first embodiment described above in the processing contents of the channel power estimation unit 65 and the filter estimation unit 66. Therefore, hereinafter, the functional configuration of the information processing apparatus 60 according to the present embodiment will be described, particularly focusing on differences from the information processing apparatus 10 according to the first embodiment described above. Detailed description of the same configuration as in FIG.

22, the channel power estimation unit 65 includes an input power estimation unit 651, a non-correlated component power estimation unit 653, and a random noise power estimation unit 655.

The input power estimation unit 651 corresponds to the channel power estimation unit 15 in the information processing apparatus 10 according to the first embodiment described above. That is, the input power estimation unit 651 is configured to output the power of each sound collecting unit 110 for each frequency based on each frequency component X _m (i, k) of the acoustic signal x _m (n) corresponding to each sound collecting unit 110. Estimate the spectrum. Then, the input power estimation unit 651 outputs the estimation result of the power spectrum P _m (i, k) of each sound collection unit 110 to the random noise power estimation unit 655 for each frequency.

The uncorrelated component power estimation unit 653 receives feedback of the output signal Y (i, k) generated by the filtering processing unit 17 performing the filtering process. Note that the output signal Y (i, k) is noise (random noise) in each frequency component X _m (i, k) of the acoustic signal x _m (n) for each sound collecting unit 110 that has been previously collected. For example, the sound of which the influence is suppressed corresponds to a frequency component of each sound collection unit 110 of a target sound such as a voice spoken by the user. Next, the decorrelation component power estimation unit 653 outputs each frequency component X _m (i, k) of the acoustic signal x _m (n) corresponding to each sound collection unit 110 and the feedback output signal Y (i, k). ) To estimate the power spectrum Q _m (i, k) of the uncorrelated component with the output signal Y (i, k). Of the frequency component X _m (i, k), the component that is uncorrelated with the output signal Y (i, k) (hereinafter, also simply referred to as “non-correlated component”) is the frequency component X _m (i, k). This corresponds to a noise component such as random noise included in k). Details of the signal processing by the uncorrelated component power estimation unit 653 will be described later. Then, the uncorrelated component power estimation unit 653 outputs the estimation result of the power spectrum Q _m (i, k) of each sound collection unit 110 to the random noise power estimation unit 655 for each frequency.

The random noise power estimation unit 655 acquires an estimation result of the power spectrum P _m (i, k) of each sound collection unit 110 for each frequency from the input power estimation unit 651. In addition, the random noise power estimation unit 655 acquires an estimation result of the power spectrum Q _m (i, k) of the uncorrelated component corresponding to each sound collecting unit 110 for each frequency from the uncorrelated component power estimation unit 653. . Then, the random noise power estimation unit 655 uses the filter estimation unit 66 to calculate the filter coefficient w (i, k) based on the obtained estimation results of the power spectra P _m (i, k) and Q _m (i, k). The power spectrum Wm (i, k) of each sound collection unit 110 for each frequency for calculation is determined. The details of the process related to the determination of the power spectrum Wm (i, k) by the random noise power estimation unit 655 will be described later. Then, the random noise power estimation unit 655 outputs information indicating the power spectrum Wm (i, k) of each sound collection unit 110 to the filter estimation unit 66 for each frequency.

Based on information indicating the power spectrum Wm (i, k) of each sound collection unit 110 for each frequency output from the channel power estimation unit 65, the filter estimation unit 66 performs a filtering process. A filter coefficient w (i, k) for calculating the value is calculated. At this time, when generating the above-described matrix R (i, k) shown as (Equation 2), the filter estimation unit 66 replaces the power spectrum P _m (i, k) with the power spectrum Wm (i , K) is different from the filter estimation unit 16 according to the first embodiment described above.

On the other hand, the filter coefficient w based on the array manifold vector a (k) and the generated matrix R (i, k) generated as described above based on the subsequent processing, that is, (Expression 3) to (Expression 6). The processing related to the calculation of (i, k) is the same as that of the filter estimation unit 16 according to the first embodiment described above. Therefore, detailed description of the contents of the processing is omitted.

As described above, the filter estimation unit 66 calculates the filter coefficient w (i, k) based on the information indicating the power spectrum Wm (i, k) of each sound collection unit 110 for each acquired frequency. The filtered filter coefficient w (i, k) is output to the filter processing unit 17. The subsequent processing is the same as that of the information processing apparatus 10 (see FIG. 6) according to the first embodiment described above.

The example of the functional configuration of the information processing apparatus according to the present embodiment has been described above with reference to FIG.

<2.3. Details of uncorrelated component power estimation unit>
Next, details of processing in which the decorrelation component power estimation unit 653 calculates the power spectrum Q _m (i, k) of the decorrelation component corresponding to each sound collection unit 110 for each frequency will be described.

First, the basic principle for the uncorrelated component power estimation unit 653 to calculate the power spectrum Q _m (i, k) will be described. The sound (signal) input to the sound collection unit such as a microphone includes, for example, a target sound S _m such as a user's voice, a so-called background noise N _m, and a random noise W _m such as a wind noise. And are included. That is, each frequency component X _m (i, k) of the acoustic signal x _m (n) for each sound collecting unit 110 is based on the target sound S _m , the background noise N _m , and the random noise W _m as follows: It is represented by the relational expression shown as 8).

... (Formula 8)

Here, the sounds (signals) input to each of the M sound collecting units are summarized as a relational expression shown as (Equation 9) below.

... (Formula 9)

In shown in (Equation 9), S is for the target sound S _m summarizes the M sound collecting unit. Similarly, N represents the background noise N _m is a summary for M sound collecting unit, W is one in which the random noise W _m summarizes the M sound collecting unit. Note that S, N, and W are each shown as a vector. S _org indicates the target sound itself output from the sound source, and is represented by a scalar value. Further, a _k corresponds to the array manifold vector a (k) described above. In other words, S indicates a component of the target sound that takes into account the effects of signal degradation, delay, and the like that occur when the target sound S _org output from the sound source propagates through the space until it reaches the sound collection unit. Yes.

Here, the generation timing of random noise W such as wind noise is random, and in the information processing apparatus according to the present disclosure, a plurality of sound collection units (particularly, sound collections arranged in a distributed manner as shown in FIG. 1). Part) can be defined as signals having no correlation.

Based on such characteristics, (Equation 9) can be defined as a relationship between vectors as shown in FIG. FIG. 23 is an explanatory diagram for explaining a basic principle of processing of the uncorrelated component power estimation unit 653. In the example shown in FIG. 23, the case where the voice uttered by the user is collected as the target sound is shown. Further, the vector space shown in FIG. 23 is defined based on the manifold vector _ak .

In FIG. 23, X represents the sound collected by the sound collection unit (that is, the input signal), and corresponds to X shown in (Equation 9). Y ideally corresponds to a component based on the estimation result of the target sound S _org with respect to the input signal X (that is, a user's utterance component). That is, the component Y schematically shows a user's speech component (or a component having a correlation with the user's speech component) among the components included in the input signal X. On the other hand, Z corresponds to a component having a small correlation (or no correlation) with the user's speech component among the components included in the input signal X.

If all of the background noise N and the random noise W can be suppressed, the component Z is only the background noise N and the random noise W. However, in the configuration in which the sound collection units are arranged around the neck as in the information processing apparatus according to the present disclosure (see, for example, FIG. 1), the background noise N is , It is observed as a component having a correlation between the sound collecting parts. Therefore, the component Y includes a background noise N component in addition to the user's utterance component S. On the other hand, random noise W such as wind noise is indicated as a component Z because the correlation with the user's speech component is small.

By utilizing the characteristics as described above, the uncorrelated component power estimation unit 653 uses the feedback of the output signal Y (that is, the user's utterance component), thereby having a small correlation (or correlation) with the output signal Y. A component having no property is extracted as a component of random noise W. In the following description, the component Z is also referred to as “non-correlated component Z”.

For example, when the number of the sound collecting units 110 is four, the array manifold vector a _k is expressed by the calculation formula shown below as (Formula 10) based on the calculation formula described above as (Formula 4).

... (Formula 10)

Here, an input signal X, based on the inner product of the manifold vector a _k, it is possible to extract a component obtained by projecting the input signal X to the manifold vectors a _k. From such characteristics, it is possible to extract the non-correlated component Z as a component orthogonal to the manifold vector _ak based on the calculation formula shown below as (Equation 11).

... (Formula 11)

Here, in (Equation 11), the component shown as (a _k ^H · a _k ) ⁻¹ · a _k ^H · X corresponds to the user's utterance component Y shown in FIG. That is, the above (Formula 11) can be expressed by a calculation formula shown as (Formula 12) below.

... (Formula 12)

Here, when the output signal Y fed back (that is, the output signal after the filtering process by the filter processing unit 17) is applied as the component Y in the above (Expression 12), the above (Expression 12) becomes the above-described (Expression 6). ) On the basis of (), it can be expressed by a calculation formula shown as (Formula 13) below.

... (Formula 13)

It is possible to estimate the power spectrum of the uncorrelated component Z by calculating the signal power based on the uncorrelated component Z calculated as described above and performing time smoothing. Here, the power spectrum Q _m (i, k) of the decorrelation component Z corresponding to the i frame and the frequency k in the m-th sound collecting unit 110 (that is, the sound collecting unit 11m) is expressed as (Equation 14) below. It is expressed by the calculation formula shown as In (Expression 14) shown below, Z _m ^* (i, k) represents a conjugate complex number of Z _m (i, k). In (Expression 14), r represents a smoothing coefficient in the frame direction for suppressing a rapid change in power spectrum (0 ≦ r <1).

... (Formula 14)

As described above, the uncorrelated component power estimation unit 653 calculates the power spectrum Q _m (i, k) of the uncorrelated component.

Note that the non-correlated component power estimation unit 653 may not necessarily collect all the sound collection units 110 if the sound collection results of two or more sound collection units 110 can be used when estimating the power spectrum Q _m (i, k). There is no need to use sound results. As a specific example, the uncorrelated component power estimation unit 653 is installed at a position where it is difficult to collect a target sound such as a voice, such as the sound collection unit 110 located behind the user's head. The sound collection result of the sound unit 110 may not be used for estimating the power spectrum Q _m (i, k).

The details of the process in which the decorrelation component power estimation unit 653 calculates the power spectrum Q _m (i, k) of the decorrelation component corresponding to each sound collection unit 110 for each frequency has been described above.

<2.4. Details of Random Noise Power Estimator>
Subsequently, details of the process in which the random noise power estimation unit 655 determines the power spectrum Wm (i, k) of each sound collection unit 110 for each frequency used for the calculation of the filter coefficient w (i, k). explain.

As described above, the random noise power estimation unit 655 includes the power spectrum P _m (i, k) acquired from the input power estimation unit 651 and the power spectrum of the non-correlation component acquired from the non-correlation component power estimation unit 653. A power spectrum Wm (i, k) is determined based on each estimation result of Q _m (i, k).

(Case of applying the power spectrum _{Q m)}
For example, the random noise power estimation unit 655 may output the estimation result of the power spectrum Q _m (i, k) of the uncorrelated component to the filter estimation unit 66 as the power spectrum Wm (i, k). In this case, the channel power estimation unit 65 may not include the input power estimation unit 651.

(Case where the power spectra P _m and Q _m are selectively switched)
As another example, the random noise power estimation unit 655 selectively selects one of the estimation results of the power spectra P _m (i, k) and Q _m (i, k) based on a predetermined condition. Alternatively, the power spectrum Wm (i, k) may be output to the filter estimation unit 66.

(Case where the power spectrum W _m is calculated adaptively)
As another example, the random noise power estimation unit 655 adaptively applies the power spectrum Wm (i, k) based on the estimation results of the power spectra P _m (i, k) and Q _m (i, k). May be calculated.

For example, the random noise power estimation unit 655 receives the power spectrums P _m (i, k) and Q _m (i, k) as inputs, and based on the calculation formula shown below as (Equation 15), the target sound (sound, etc.) The power spectrum W _m ^˜ taking into account the relationship between the noise and the random noise is calculated. Note that “W _m ^˜ ” indicates characters with a tilde on “W _m ”. Further, Pm and Qm shown below are generalized descriptions of the power spectra P _m (i, k) and Q _m (i, k).

... (Formula 15)

For example, the following (Equation 16) calculates the power spectrum W _m ^˜ taking into account the relationship between the target sound and random noise, using the power spectra P _m (i, k) and Q _m (i, k) as inputs. A specific example of the function F for doing this is shown.

... (Formula 16)

Then, the random noise power estimation unit 655 calculates the power spectrum Wm based on the calculation formula shown below as (Equation 17) based on the power spectrum W _m ^˜ in consideration of the relationship between the target sound and the random noise. . In (Equation 17), r represents a smoothing coefficient in the frame direction for suppressing a sudden change in power spectrum (0 ≦ r <1). That is, the random noise power estimation unit 655 may smooth the power spectrum Wm calculated based on the calculation formula shown below as (Equation 17) between frames based on the setting of the coefficient r.

... (Formula 17)

Here, the power spectrum P _m shown in (Equation 16), that is, the estimation result of the power spectrum P _m (i, k) by the input power estimation unit 651 is collected by the sound collection unit 110 as described above. Corresponds to the level of sound. On the other hand, the power spectrum Q _m shown in (Equation 16), that is, the estimation result of the power spectrum Q _m (i, k) by the uncorrelated component power estimation unit 653 is the random noise such as wind noise. Corresponds to the level. That is, the weight Q _m / (P _m + Q _m ) shown in (Equation 16) changes based on the relationship between the target sound such as speech and random noise such as wind noise.

Specifically, when the signal level of the target sound is sufficiently large with respect to random noise, the influence of the power spectrum P _m becomes dominant, and the weight Q _m / (P _m + Q _m ) becomes smaller. That is, the weight Q _m / (P _m + Q _m ) in this case indicates control that further suppresses use of the sound collection result of the corresponding channel (that is, the sound collection unit 110). Here, the reciprocal of the weight Q _m / (P _m + Q _m ) is applied to the calculation of the filter coefficient w (i, k). Therefore, when the signal level of the target sound is sufficiently high with respect to random noise, the filter coefficient w (i, k) is calculated so that the use of the sound collection result by the corresponding channel is given higher priority. It becomes.

On the other hand, is greater than the effect of random noise such as wind noise, the influence of the power spectrum Q _m becomes more dominant, weight _{_{Q m / (P m + Q}} m) is greater. That is, the weight Q _m / (P _m + Q _m ) in this case indicates control that gives priority to the use of the sound collection result of the corresponding channel (that is, the sound collection unit 110). As described above, the reciprocal of the weight Q _m / (P _m + Q _m ) is applied to the calculation of the filter coefficient w (i, k). Therefore, when the influence of random noise is greater, the filter coefficient w (i, k) is calculated so that the use of the sound collection result by the corresponding channel is suppressed.

In other words, under the circumstances where random noise such as wind noise is small and the sound is mainly collected by the control described above, the sound collecting unit 110 that has collected the sound at a higher sound level. The result is used more preferentially to calculate the filter coefficient w (i, k). On the other hand, under the situation where the influence of random noise such as wind noise is large, the sound collection result of the sound collection unit 110 having a smaller observation level is more preferentially the same as in the first embodiment described above. Used to calculate the filter coefficient w (i, k). As described above, the random noise power estimation unit 655 uses the power spectrum Wm (i, k) for calculating the filter coefficient w (i, k) as the target sound such as speech and the wind noise. It is possible to calculate adaptively according to the relationship with random noise.

Then, the random noise power estimation unit 655 may output the power spectrum Wm (i, k) calculated based on (Equation 17) to the filter estimation unit 66.

The details of the process in which the random noise power estimation unit 655 determines the power spectrum Wm (i, k) of each sound collection unit 110 for each frequency used for calculating the filter coefficient w (i, k) have been described above. did. The example described above is merely an example, and the power spectrum Wm (i, k) is calculated based on the estimation result of at least one of the power spectra P _m (i, k) and Q _m (i, k). The content is not particularly limited as long as it can be determined.

<2.5. Evaluation>
As described above, the information processing apparatus 60 according to this embodiment includes the sound collection results obtained by at least two or more sound collection units 110 among the plurality of sound collection units 110 and the output signal Y (i, The power spectrum Q _m (i, k) of the non-correlated component is estimated based on the feedback of k). The information processing apparatus 60 uses the estimation result of the power spectrum Q _m (i, k) of the non-correlated component for estimation of the filter coefficient w (i, k). With such a configuration, the information processing apparatus 60 maintains the effect of suppressing noise generated randomly such as wind noise as in the first embodiment described above, and further, the influence of noise generated randomly is small. In some cases, the target sound can be acquired in a more preferable manner.

In the above description, the signal processing according to the present embodiment has been described focusing on the case where it is applied to the so-called neckband type wearable device shown in FIG. On the other hand, the application destination of the signal processing according to the present embodiment is not necessarily limited to the example shown in FIG. Specifically, the signal processing according to the present embodiment can be applied to any device that includes a plurality of sound collection units. More preferably, the plurality of sound collection units may be arranged so that the distances from the sound source of the target sound (for example, the mouth where the voice is spoken) are different from each other. More preferably, the plurality of sound collecting units may be arranged so as to be positioned in different directions with respect to the sound source of the target sound.

<< 3. Hardware configuration >>
Next, an example of a hardware configuration of the information processing apparatus 10 (that is, the above-described signal processing apparatuses 11 to 14) according to each embodiment of the present disclosure will be described with reference to FIG. FIG. 24 is a diagram illustrating an example of a hardware configuration of the information processing apparatus 10 according to each embodiment of the present disclosure.

As illustrated in FIG. 24, the information processing apparatus 10 according to the present embodiment includes a processor 901, a memory 903, a storage 905, an operation device 907, a notification device 909, an acoustic device 911, and a sound collection device 913. And bus 917. Further, the information processing apparatus 10 may include a communication device 915.

The processor 901 may be, for example, a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a DSP (Digital Signal Processor), or a SoC (System on Chip), and executes various processes of the information processing apparatus 10. The processor 901 can be configured by, for example, an electronic circuit for executing various arithmetic processes. Note that the frequency resolving unit 13, the channel power estimating unit 15, the filter estimating unit 16, the filter processing unit 17, and the frequency synthesizing unit 18 described above can be realized by the processor 901.

The memory 903 includes RAM (Random Access Memory) and ROM (Read Only Memory), and stores programs and data executed by the processor 901. The storage 905 can include a storage medium such as a semiconductor memory or a hard disk.

The operation device 907 has a function of generating an input signal for a user to perform a desired operation. The operation device 907 can be configured as a touch panel, for example. As another example, the operation device 907 generates an input signal based on an input by the user, such as buttons, switches, and a keyboard, and an input for the user to input information, and supplies the input signal to the processor 901. It may be composed of a control circuit or the like.

The notification device 909 is an example of an output device, and may be a device such as a liquid crystal display (LCD) device or an organic EL (OLED: Organic Light Emitting Diode) display, for example. In this case, the notification device 909 can notify the user of predetermined information by displaying the screen.

In addition, the example of the notification device 909 described above is merely an example, and the aspect of the notification device 909 is not particularly limited as long as predetermined information can be notified to the user. As a specific example, the notification device 909 may be a device that notifies the user of predetermined information using a lighting or blinking pattern, such as an LED (Light Emitting Diode). Further, the notification device 909 may be a device that notifies a user of predetermined information by vibrating like a so-called vibrator.

The acoustic device 911 is a device that notifies a user of predetermined information by outputting a predetermined acoustic signal, such as a speaker.

The sound collection device 913 is a device such as a microphone that collects sound emitted from the user and the sound of the surrounding environment and acquires it as acoustic information (acoustic signal). In addition, the sound collection device 913 may acquire data indicating an analog sound signal indicating collected sound or sound as sound information, or convert the analog sound signal into a digital sound signal, Data indicating a later digital acoustic signal may be acquired as acoustic information. Note that the above-described sound collection unit 110 (for example, the sound collection units 111 to 11M shown in FIG. 6) can be realized by the sound collection device 913.

The communication device 915 is a communication unit included in the information processing apparatus 10 and communicates with an external apparatus via a network. The communication device 915 is a wired or wireless communication interface. When the communication device 915 is configured as a wireless communication interface, the communication device 915 may include a communication antenna, an RF (Radio Frequency) circuit, a baseband processor, and the like.

The communication device 915 has a function of performing various kinds of signal processing on a signal received from an external device, and can supply a digital signal generated from the received analog signal to the processor 901.

The bus 917 connects the processor 901, the memory 903, the storage 905, the operation device 907, the notification device 909, the acoustic device 911, the sound collection device 913, and the communication device 915 to each other. The bus 917 may include a plurality of types of buses.

In addition, it is possible to create a program for causing hardware such as a processor, memory, and storage built in a computer to exhibit functions equivalent to the configuration of the information processing apparatus 10 described above. A computer-readable storage medium that records the program can also be provided.

<< 4. Conclusion >>
As described above, the information processing apparatus 10 according to the present embodiment includes a convex portion having a streamline shape at least in part, and is positioned at or near the tip of the convex portion. The sound collection unit 110 is supported. By adopting such a configuration, for example, the effect of random noise such as wind noise, noise associated with vibration, and rubbing caused by wearing of the device is reduced, and the target sound ( For example, the user's voice) can be collected.

Moreover, the information processing apparatus 10 according to the present embodiment may include a plurality of sound collecting units 110 and may be supported so that the plurality of sound collecting units 110 face different directions. With such a configuration, even in a situation where noise such as wind noise, vibration, and rubbing due to wearing of the device occurs randomly, some sound collection units (i.e., the influence of noise is Based on the sound collection results of a small number of sound collection units), it becomes possible to supplement the characteristics of other sound collection units.

The preferred embodiments of the present disclosure have been described in detail above with reference to the accompanying drawings, but the technical scope of the present disclosure is not limited to such examples. It is obvious that a person having ordinary knowledge in the technical field of the present disclosure can come up with various changes or modifications within the scope of the technical idea described in the claims. Of course, it is understood that it belongs to the technical scope of the present disclosure.

In addition, the effects described in this specification are merely illustrative or illustrative, and are not limited. That is, the technology according to the present disclosure can exhibit other effects that are apparent to those skilled in the art from the description of the present specification in addition to or instead of the above effects.

The following configurations also belong to the technical scope of the present disclosure.
(1)
The sound collection section;
A support member for supporting the sound collection unit so as to be located at or near the tip of the protrusion, or at least part of the protrusion having a streamlined shape;
An information processing apparatus comprising:
(2)
The information processing apparatus according to (1), further including one or more second sound collecting units different from the first sound collecting unit in addition to the first sound collecting unit which is the sound collecting unit.
(3)
The information processing apparatus according to (2), wherein the support member supports each of the plurality of second sound collection units so as to face different directions.
(4)
The information processing apparatus according to (1), wherein the support member is attached to a predetermined part of a user and supports the sound collection unit so that the sound collection unit and the part have a predetermined positional relationship.
(5)
The part is a neck;
When the support member is attached to the neck, the convex portion is provided so that the tip of the convex portion faces substantially forward of the user.
The information processing apparatus according to (4).
(6)
In addition to the first sound collection unit that is the sound collection unit, a plurality of second sound collection units different from the first sound collection unit are provided,
Supporting at least two or more second sound collecting parts among the plurality of second sound collecting parts at positions that are substantially symmetrical with each other with respect to the part;
The information processing apparatus according to (4) or (5).
(7)
Based on the sound collected by each of the first sound collection unit and the one or more second sound collection units, a noise component for sound coming from a predetermined direction with respect to the first sound collection unit The information processing apparatus according to (2), further including a signal processing unit that suppresses noise.
(8)
The signal processing unit estimates a signal level of each frequency component of the sound based on the sound collected by each of the first sound collecting unit and the one or more second sound collecting units, and The information processing apparatus according to (7), wherein the noise component is suppressed based on a signal level estimation result.
(9)
The signal processing unit includes a first sound collected by each of at least a plurality of sound collection units of the first sound collection unit and the one or more second sound collection units, and a previous process. The information processing apparatus according to (7), wherein the noise component included in the first sound is suppressed based on a correlation with the second sound in which the noise component is suppressed.
(10)
The support member supports the plurality of sound collection units such that a distance between each of at least two or more of the plurality of sound collection units and a predetermined sound source is different from each other. The information processing apparatus according to 9).
(11)
The support member supports the plurality of sound collection units such that at least two or more of the plurality of sound collection units are located in different directions with respect to a predetermined sound source, (9) The information processing apparatus according to (10).
(12)
The support member is a housing having a substantially rectangular surface at least in part,
The housing has the convex portion in a predetermined region including a corner of the substantially rectangular surface, and supports the sound collecting portion at the tip of the convex portion or in the vicinity of the tip.
The information processing apparatus according to (1).
(13)
A plurality of the sound collecting units;
The housing has the convex portion in a predetermined region including the corner for each of a plurality of corners of the substantially rectangular surface, and at or near the tip of the convex portion. The information processing apparatus according to (12), which supports the sound collection unit.
(14)
A band portion for supporting the housing with respect to a user's arm;
When the band portion is attached to the arm, the band portion includes another sound collecting portion different from the sound collecting portion at a position that is substantially symmetrical with the housing with respect to the arm.
The information processing apparatus according to (12) or (13).
(15)
The support member is a glasses-type frame attached to the user's head,
The frame has the convex portion on at least a part of the front, and supports the sound collecting portion at the tip of the convex portion or in the vicinity of the tip.
The information processing apparatus according to (1).
(16)
The information processing according to (15), wherein the frame has the convex portion in the vicinity of the bridge or the bridge, and supports the sound collecting portion in a front end of the convex portion or in the vicinity of the front end. apparatus.

DESCRIPTION OF SYMBOLS 10 Information processing apparatus 13 Frequency decomposition part 15 Channel power estimation part 16 Filter estimation part 17 Filter processing part 18 Frequency synthesis part 110-113 Sound collection part 60 Information processing apparatus 65 Channel power estimation part 651 Input power estimation part 653 Uncorrelated component power Estimator 655 Random noise power estimator 66 Filter estimator

Claims

The sound collection section;
A support member for supporting the sound collection unit so as to be located at or near the tip of the protrusion, or at least part of the protrusion having a streamlined shape;
An information processing apparatus comprising:
The information processing apparatus according to claim 1, further comprising one or more second sound collecting units different from the first sound collecting unit in addition to the first sound collecting unit which is the sound collecting unit.
The information processing apparatus according to claim 2, wherein the support member supports each of the plurality of second sound collection units so as to face different directions.
The information processing apparatus according to claim 1, wherein the support member is attached to a predetermined part of a user and supports the sound collection unit so that the sound collection unit and the part have a predetermined positional relationship.
The part is a neck;
When the support member is attached to the neck, the convex portion is provided so that the tip of the convex portion faces substantially forward of the user.
The information processing apparatus according to claim 4.
In addition to the first sound collection unit that is the sound collection unit, a plurality of second sound collection units different from the first sound collection unit are provided,
Supporting at least two or more second sound collecting parts among the plurality of second sound collecting parts at positions that are substantially symmetrical with each other with respect to the part;
The information processing apparatus according to claim 4.
Based on the sound collected by each of the first sound collection unit and the one or more second sound collection units, a noise component for sound coming from a predetermined direction with respect to the first sound collection unit The information processing apparatus according to claim 2, further comprising: a signal processing unit that suppresses noise.
The signal processing unit estimates a signal level of each frequency component of the sound based on the sound collected by each of the first sound collecting unit and the one or more second sound collecting units, and The information processing apparatus according to claim 7, wherein the noise component is suppressed based on a signal level estimation result.
The signal processing unit includes a first sound collected by each of at least a plurality of sound collection units of the first sound collection unit and the one or more second sound collection units, and a previous process. The information processing apparatus according to claim 7, wherein the noise component included in the first sound is suppressed based on a correlation with the second sound in which the noise component is suppressed.
The said support member supports the said several sound collection part so that the distance between each of at least 2 or more sound collection parts of the said several sound collection parts and a predetermined sound source may mutually differ. 9. The information processing apparatus according to 9.
The support member supports the plurality of sound collection units such that at least two or more of the plurality of sound collection units are located in different directions with respect to a predetermined sound source. Item 10. The information processing device according to Item 9.
The support member is a housing having a substantially rectangular surface at least in part,
The housing has the convex portion in a predetermined region including a corner of the substantially rectangular surface, and supports the sound collecting portion at the tip of the convex portion or in the vicinity of the tip.
The information processing apparatus according to claim 1.
A plurality of the sound collecting units;
The housing has the convex portion in a predetermined region including the corner for each of a plurality of corners of the substantially rectangular surface, and at or near the tip of the convex portion. The information processing apparatus according to claim 12, wherein the information collection unit is supported.
A band portion for supporting the housing with respect to a user's arm;
When the band portion is attached to the arm, the band portion includes another sound collecting portion different from the sound collecting portion at a position that is substantially symmetrical with the housing with respect to the arm.
The information processing apparatus according to claim 12.
The support member is a glasses-type frame attached to the user's head,
The frame has the convex portion on at least a part of the front, and supports the sound collecting portion at the tip of the convex portion or in the vicinity of the tip.
The information processing apparatus according to claim 1.
The information processing apparatus according to claim 15, wherein the frame has the convex portion in the vicinity of the bridge or the bridge, and supports the sound collection unit in a front end of the convex portion or in the vicinity of the front end. .