Patents by Inventor Kohhei Hayashida
Kohhei Hayashida has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).
-
Patent number: 11238852Abstract: A speech translation device includes: a beamformer which forms, from a speech signal obtained by a microphone array, a first beam and a second beam having different directions; a direction designator which designates one of the first beam and the second beam according to a user operation; a signal-to-noise (SN) ratio calculator which calculates an SN ratio using the designated beam as a signal component in the SN ratio, and the other beam not designated as a noise component; a display determiner which determines whether recognition of the designated beam is difficult, using the calculated SN ratio, and determines a speaking instruction for overcoming difficulty of the recognition when determining that the recognition is difficult; and a display which displays the speaking instruction determined by the display determiner in a display area.Type: GrantFiled: March 25, 2019Date of Patent: February 1, 2022Assignee: PANASONIC CORPORATIONInventors: Kohhei Hayashida, Atsushi Sakaguchi, Kazuya Nomura, Hiroki Furukawa, Tsuyoki Nishikawa
-
Patent number: 11182567Abstract: A speech translation apparatus includes: an estimator which estimates a sound source direction, based on an acoustic signal obtained by a microphone array unit; a controller which identifies that an utterer is a user or a conversation partner, based on the sound source direction estimated after the start of translation is instructed by a button, using a positional relationship indicated by a layout information item stored in storage and selected in advance, and determines a translation direction indicating input and output languages in and into which content of the acoustic signal is recognized and translated, respectively; and a translator which obtains, according to the translation direction, original text indicating the content in the input language and translated text indicating the content in the output language. The controller displays the original and translated texts on first and second display areas corresponding to the positions of the user and conversation partner, respectively.Type: GrantFiled: March 26, 2019Date of Patent: November 23, 2021Assignee: PANASONIC CORPORATIONInventors: Hiroki Furukawa, Tsuyoki Nishikawa, Atsushi Sakaguchi, Kohhei Hayashida
-
Patent number: 10524051Abstract: A sound source direction estimation device includes: a phase difference calculator which calculates, from an acoustic signal obtained by a microphone array, a first phase difference of a pair of microphone units; a similarity calculator which calculates similarities between the calculated first phase difference and second phase differences precalculated for directions and stored in a phase difference database; a peak searcher which searches for a direction for which a highest similarity is calculated by the similarity calculator, and estimates the direction searched out to be a sound source direction; a feature quantity calculator which uses the calculated similarities, the estimated sound source direction, and an acoustic feature quantity obtained from the obtained acoustic signal, to calculate a feature quantity obtained by correcting the acoustic feature quantity; and a speech/non-speech determiner which determines whether the obtained acoustic signal indicates speech, using the feature quantity calculatedType: GrantFiled: March 28, 2019Date of Patent: December 31, 2019Assignee: PANASONIC CORPORATIONInventors: Kohhei Hayashida, Takeo Kanamori
-
Publication number: 20190306620Abstract: A sound source direction estimation device includes: a phase difference calculator which calculates, from an acoustic signal obtained by a microphone array, a first phase difference of a pair of microphone units; a similarity calculator which calculates similarities between the calculated first phase difference and second phase differences precalculated for directions and stored in a phase difference database; a peak searcher which searches for a direction for which a highest similarity is calculated by the similarity calculator, and estimates the direction searched out to be a sound source direction; a feature quantity calculator which uses the calculated similarities, the estimated sound source direction, and an acoustic feature quantity obtained from the obtained acoustic signal, to calculate a feature quantity obtained by correcting the acoustic feature quantity; and a speech/non-speech determiner which determines whether the obtained acoustic signal indicates speech, using the feature quantity calculatedType: ApplicationFiled: March 28, 2019Publication date: October 3, 2019Inventors: Kohhei HAYASHIDA, Takeo KANAMORI
-
Publication number: 20190304442Abstract: A speech translation device includes: a beamformer which forms, from a speech signal obtained by a microphone array, a first beam and a second beam having different directions; a direction designator which designates one of the first beam and the second beam according to a user operation; a signal-to-noise (SN) ratio calculator which calculates an SN ratio using the designated beam as a signal component in the SN ratio, and the other beam not designated as a noise component; a display determiner which determines whether recognition of the designated beam is difficult, using the calculated SN ratio, and determines a speaking instruction for overcoming difficulty of the recognition when determining that the recognition is difficult; and a display which displays the speaking instruction determined by the display determiner in a display area.Type: ApplicationFiled: March 25, 2019Publication date: October 3, 2019Inventors: Kohhei HAYASHIDA, Atsushi SAKAGUCHI, Kazuya NOMURA, Hiroki FURUKAWA, Tsuyoki NISHIKAWA
-
Publication number: 20190303443Abstract: A speech translation apparatus includes: an estimator which estimates a sound source direction, based on an acoustic signal obtained by a microphone array unit; a controller which identifies that an utterer is a user or a conversation partner, based on the sound source direction estimated after the start of translation is instructed by a button, using a positional relationship indicated by a layout information item stored in storage and selected in advance, and determines a translation direction indicating input and output languages in and into which content of the acoustic signal is recognized and translated, respectively; and a translator which obtains, according to the translation direction, original text indicating the content in the input language and translated text indicating the content in the output language. The controller displays the original and translated texts on first and second display areas corresponding to the positions of the user and conversation partner, respectively.Type: ApplicationFiled: March 26, 2019Publication date: October 3, 2019Inventors: Hiroki FURUKAWA, Tsuyoki NISHIKAWA, Atsushi SAKAGUCHI, Kohhei HAYASHIDA
-
Patent number: 10395644Abstract: A speech recognition method acquires sound information via multiple microphones, detects a sound source interval including sound from the sound information, acquires an estimated direction of speech by conducting direction estimation on a speech interval from among the sound source interval, conducts an adaptation process of using the sound information to estimate filter coefficients, decides a buffer size of the sound information to hold in a buffer, based on sound source interval information, estimated direction information, and adaptation process convergence state information, holds the sound information in the buffer according to the buffer size, conducts a beamforming process using the sound information held in the buffer and the filter coefficients to acquire speech information, and conducts speech recognition on the speech information acquired by the beamforming process.Type: GrantFiled: February 13, 2017Date of Patent: August 27, 2019Assignee: PANASONIC CORPORATIONInventors: Kohhei Hayashida, Tsuyoki Nishikawa
-
Patent number: 10360922Abstract: A noise reduction device includes an acoustic characteristic determination unit, a microphone pair selection unit, and a noise reduction unit. The acoustic characteristic determination unit determines, using one of microphone pair signals obtained from a plurality of microphone pairs, whether acoustic characteristics of each of the plurality of microphone pairs satisfy a certain requirement. The microphone pair selection unit selects, from among the plurality of microphone pairs, a target microphone pair whose acoustic characteristics have been determined to satisfy the certain requirement. The noise reduction unit reduces, using a microphone pair signal obtained from the target microphone pair, noise included in an input signal obtained from at least one of microphone signals output from a plurality of microphones.Type: GrantFiled: September 13, 2017Date of Patent: July 23, 2019Assignee: PANASONIC CORPORATIONInventors: Tsuyoki Nishikawa, Kohhei Hayashida
-
Patent number: 10356508Abstract: A sound collecting apparatus capable of effectively suppressing sounds other than a target sound is provided. The sound collecting apparatus includes a plurality of microphones. A total number of effective microphone pairs in which a distance between two microphones is smaller than a distance D is larger than a total number of the plurality of microphones. The distance D is represented by D=c/2f, where the frequency of the target sound acquired from each of the plurality of microphones is f and sound velocity is c. When an angle formed by a straight line connecting two microphones configuring an effective microphone pair and a predetermined straight line is ?, the angles ? of all effective microphone pairs acquired from the plurality of microphones are different from each other.Type: GrantFiled: June 5, 2018Date of Patent: July 16, 2019Assignee: PANASONIC INTELLECTUAL PROPERTY CORPORATION OF AMERICAInventors: Kohhei Hayashida, Tsuyoki Nishikawa, Takeo Kanamori
-
Patent number: 10264350Abstract: A sound source probing apparatus, including storage and processing circuitry, is provided that probes a direction of a sound source. The processing circuitry performs operations including determining a first correlation matrix that is a correlation matrix of acoustic signals acquired as observation signals by a microphone array including two or more microphones disposed apart from each other. The operations also include determining, by learning, weights such that a linear sum of a plurality of second correlation matrices multiplied by the respective weights is equal to the first correlation matrix where the plurality of second correlation matrices are correlation matrices, which are determined for respective directions determined based on an array arrangement of the microphone array and which are stored in advance in the storage.Type: GrantFiled: February 19, 2018Date of Patent: April 16, 2019Assignee: PANASONIC INTELLECTUAL PROPERTY CORPORATION OF AMERICAInventors: Takeo Kanamori, Kohhei Hayashida, Shintaro Yoshikuni
-
Publication number: 20180376239Abstract: A sound collecting apparatus capable of effectively suppressing sounds other than a target sound is provided. The sound collecting apparatus includes a plurality of microphones. A total number of effective microphone pairs in which a distance between two microphones is smaller than a distance D is larger than a total number of the plurality of microphones. The distance D is represented by D=c/2f, where the frequency of the target sound acquired from each of the plurality of microphones is f and sound velocity is c. When an angle formed by a straight line connecting two microphones configuring an effective microphone pair and a predetermined straight line is ?, the angles ? of all effective microphone pairs acquired from the plurality of microphones are different from each other.Type: ApplicationFiled: June 5, 2018Publication date: December 27, 2018Inventors: KOHHEI HAYASHIDA, TSUYOKI NISHIKAWA, TAKEO KANAMORI
-
Publication number: 20180255395Abstract: A sound source probing apparatus, including storage and processing circuitry, is provided that probes a direction of a sound source. The processing circuitry performs operations including determining a first correlation matrix that is a correlation matrix of acoustic signals acquired as observation signals by a microphone array including two or more microphones disposed apart from each other. The operations also include determining, by learning, weights such that a linear sum of a plurality of second correlation matrices multiplied by the respective weights is equal to the first correlation matrix where the plurality of second correlation matrices are correlation matrices, which are determined for respective directions determined based on an array arrangement of the microphone array and which are stored in advance in the storage.Type: ApplicationFiled: February 19, 2018Publication date: September 6, 2018Inventors: TAKEO KANAMORI, KOHHEI HAYASHIDA, SHINTARO YOSHIKUNI
-
Publication number: 20180096697Abstract: A noise reduction device includes an acoustic characteristic determination unit, a microphone pair selection unit, and a noise reduction unit. The acoustic characteristic determination unit determines, using one of microphone pair signals obtained from a plurality of microphone pairs, whether acoustic characteristics of each of the plurality of microphone pairs satisfy a certain requirement. The microphone pair selection unit selects, from among the plurality of microphone pairs, a target microphone pair whose acoustic characteristics have been determined to satisfy the certain requirement. The noise reduction unit reduces, using a microphone pair signal obtained from the target microphone pair, noise included in an input signal obtained from at least one of microphone signals output from a plurality of microphones.Type: ApplicationFiled: September 13, 2017Publication date: April 5, 2018Inventors: TSUYOKI NISHIKAWA, KOHHEI HAYASHIDA
-
Patent number: 9820043Abstract: A first correlation matrix, which corresponds to observed signals, is calculated, a non-scan range is specified, a second correlation matrix, which corresponds to an acoustic signal from a sound source within the non-scan range, is estimated, a third correlation matrix, which corresponds to a target sound source within a scan range, is calculated by removing the second correlation matrix from the first correlation matrix, and a first spatial spectrum, which is a localization result, is calculated from the third correlation matrix. In the estimation, the second correlation matrix is estimated from direction vectors calculated from a direction range of the non-scan range and a second spatial spectrum, which is a localization result calculated immediately before the first spatial spectrum is calculated.Type: GrantFiled: February 13, 2017Date of Patent: November 14, 2017Assignee: PANASONIC INTELLECTUAL PROPERTY CORPORATION OF AMERICAInventors: Takeo Kanamori, Kohhei Hayashida, Shintaro Yoshikuni
-
Publication number: 20170249936Abstract: A speech recognition method acquires sound information via multiple microphones, detects a sound source interval including sound from the sound information, acquires an estimated direction of speech by conducting direction estimation on a speech interval from among the sound source interval, conducts an adaptation process of using the sound information to estimate filter coefficients, decides a buffer size of the sound information to hold in a buffer, based on sound source interval information, estimated direction information, and adaptation process convergence state information, holds the sound information in the buffer according to the buffer size, conducts a beamforming process using the sound information held in the buffer and the filter coefficients to acquire speech information, and conducts speech recognition on the speech information acquired by the beamforming process.Type: ApplicationFiled: February 13, 2017Publication date: August 31, 2017Inventors: Kohhei Hayashida, Tsuyoki Nishikawa
-
Publication number: 20170251300Abstract: A first correlation matrix, which corresponds to observed signals, is calculated, a non-scan range is specified, a second correlation matrix, which corresponds to an acoustic signal from a sound source within the non-scan range, is estimated, a third correlation matrix, which corresponds to a target sound source within a scan range, is calculated by removing the second correlation matrix from the first correlation matrix, and a first spatial spectrum, which is a localization result, is calculated from the third correlation matrix. In the estimation, the second correlation matrix is estimated from direction vectors calculated from a direction range of the non-scan range and a second spatial spectrum, which is a localization result calculated immediately before the first spatial spectrum is calculated.Type: ApplicationFiled: February 13, 2017Publication date: August 31, 2017Inventors: TAKEO KANAMORI, KOHHEI HAYASHIDA, SHINTARO YOSHIKUNI