Patents by Inventor Andrew Paul Breen
Andrew Paul Breen 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).
-
Publication number: 20240129680Abstract: A hearing assistance system is disclosed that comprises a hearing assistance device configured to reproduce sounds and to assist a person to hear the sounds, the hearing assistance device including a left hearing assistance device and a right hearing assistance device, and a computing device in communication with the hearing assistance device. The computing device interacts with the hearing assistance device to implement a fitting mode wherein the hearing assistance device is caused to generate fitting sounds usable to evaluate whether the left and right hearing devices are properly fitted into respective left and right ears of the person. The computing device produces a communication indicative of whether the left and right hearing assistance devices are properly fitted into respective left and right ears of the person.Type: ApplicationFiled: December 26, 2023Publication date: April 18, 2024Applicant: NUHEARA IP PTY LTDInventors: Clint MATHURINE, Andrew Victor CAMMELL, Gregory Paul BREEN, Peng JIANG, David Ronald WARD, Alan DAVIS
-
Publication number: 20240013770Abstract: During text-to-speech processing, a speech model creates output audio data, including speech, that corresponds to input text data that includes a representation of the speech. A spectrogram estimator estimates a frequency spectrogram of the speech; the corresponding frequency-spectrogram data is used to condition the speech model. A plurality of acoustic features corresponding to different segments of the input text data, such as phonemes, syllable-level features, and/or word-level features, may be separately encoded into context vectors; the spectrogram estimator uses these separate context vectors to create the frequency spectrogram.Type: ApplicationFiled: June 6, 2023Publication date: January 11, 2024Inventors: Jaime Lorenzo Trueba, Thomas Renaud Drugman, Viacheslav Klimkov, Srikanth Ronanki, Thomas Edward Merritt, Andrew Paul Breen, Roberto Barra-Chicote
-
Patent number: 11823655Abstract: A speech-processing system receives both text data and natural-understanding data (e.g., a domain, intent, and/or entity) related to a command represented in the text data. The system uses the natural-understanding data to vary vocal characteristics in determining spectrogram data corresponding to the text data based on the natural-understanding data.Type: GrantFiled: June 9, 2022Date of Patent: November 21, 2023Assignee: Amazon Technologies, Inc.Inventors: Antonio Bonafonte, Panagiotis Agis Oikonomou Filandras, Bartosz Perz, Arent van Korlaar, Ioannis Douratsos, Jonas Felix Ananda Rohnke, Elena Sokolova, Andrew Paul Breen, Nikhil Sharma
-
Patent number: 11763797Abstract: A speech model includes a sub-model corresponding to a vocal attribute. The speech model generates an output waveform using a sample model, which receives text data, and a conditioning model, which receives text metadata and produces a prosody output for use by the sample model. If, during training or runtime, a different vocal attribute is desired or needed, the sub-model is re-trained or switched to a different sub-model corresponding to the different vocal attribute.Type: GrantFiled: June 23, 2020Date of Patent: September 19, 2023Assignee: Amazon Technologies, Inc.Inventors: Roberto Barra Chicote, Adam Franciszek Nadolski, Thomas Edward Merritt, Bartosz Putrycz, Andrew Paul Breen
-
Patent number: 11735162Abstract: During text-to-speech processing, a speech model creates output audio data, including speech, that corresponds to input text data that includes a representation of the speech. A spectrogram estimator estimates a frequency spectrogram of the speech; the corresponding frequency-spectrogram data is used to condition the speech model. A plurality of acoustic features corresponding to different segments of the input text data, such as phonemes, syllable-level features, and/or word-level features, may be separately encoded into context vectors; the spectrogram estimator uses these separate context vectors to create the frequency spectrogram.Type: GrantFiled: August 8, 2022Date of Patent: August 22, 2023Assignee: Amazon Technologies, Inc.Inventors: Jaime Lorenzo Trueba, Thomas Renaud Drugman, Viacheslav Klimkov, Srikanth Ronanki, Thomas Edward Merritt, Andrew Paul Breen, Roberto Barra-Chicote
-
Publication number: 20230113297Abstract: A speech-processing system receives both text data and natural-understanding data (e.g., a domain, intent, and/or entity) related to a command represented in the text data. The system uses the natural-understanding data to vary vocal characteristics in determining spectrogram data corresponding to the text data based on the natural-understanding data.Type: ApplicationFiled: June 9, 2022Publication date: April 13, 2023Inventors: Antonio Bonafonte, Panagiotis Agis Oikonomou Filandras, Bartosz Perz, Arent van Korlaar, Ioannis Douratsos, Jonas Felix Ananda Rohnke, Elena Sokolova, Andrew Paul Breen, Nikhil Sharma
-
Publication number: 20230058658Abstract: During text-to-speech processing, a speech model creates output audio data, including speech, that corresponds to input text data that includes a representation of the speech. A spectrogram estimator estimates a frequency spectrogram of the speech; the corresponding frequency-spectrogram data is used to condition the speech model. A plurality of acoustic features corresponding to different segments of the input text data, such as phonemes, syllable-level features, and/or word-level features, may be separately encoded into context vectors; the spectrogram estimator uses these separate context vectors to create the frequency spectrogram.Type: ApplicationFiled: August 8, 2022Publication date: February 23, 2023Inventors: Jaime Lorenzo Trueba, Thomas Renaud Drugman, Viacheslav Klimkov, Srikanth Ronanki, Thomas Edward Merritt, Andrew Paul Breen, Roberto Barra-Chicote
-
Publication number: 20230043916Abstract: During text-to-speech processing, a speech model creates synthesized speech that corresponds to input data. The speech model may include an encoder for encoding the input data into a context vector and a decoder for decoding the context vector into spectrogram data. The speech model may further include a voice decoder that receives vocal characteristic data representing a desired vocal characteristic of synthesized speech. The voice decoder may process the vocal characteristic data to determine configuration data, such as weights, for use by the speech decoder.Type: ApplicationFiled: June 24, 2022Publication date: February 9, 2023Inventors: Roberto Barra Chicote, Vatsal Aggarwal, Andrew Paul Breen, Javier Gonzalez Hernandez, Nishant Prateek
-
Patent number: 11545134Abstract: Techniques for the generation of dubbed audio for an audio/video are described.Type: GrantFiled: December 10, 2019Date of Patent: January 3, 2023Assignee: Amazon Technologies, Inc.Inventors: Marcello Federico, Robert Enyedi, Yaser Al-Onaizan, Roberto Barra-Chicote, Andrew Paul Breen, Ritwik Giri, Mehmet Umut Isik, Arvindh Krishnaswamy, Hassan Sawaf
-
Patent number: 11410639Abstract: During text-to-speech processing, a speech model creates output audio data, including speech, that corresponds to input text data that includes a representation of the speech. A spectrogram estimator estimates a frequency spectrogram of the speech; the corresponding frequency-spectrogram data is used to condition the speech model. A plurality of acoustic features corresponding to different segments of the input text data, such as phonemes, syllable-level features, and/or word-level features, may be separately encoded into context vectors; the spectrogram estimator uses these separate context vectors to create the frequency spectrogram.Type: GrantFiled: July 7, 2020Date of Patent: August 9, 2022Assignee: Amazon Technologies, Inc.Inventors: Jaime Lorenzo Trueba, Thomas Renaud Drugman, Viacheslav Klimkov, Srikanth Ronanki, Thomas Edward Merritt, Andrew Paul Breen, Roberto Barra-Chicote
-
Patent number: 11373633Abstract: During text-to-speech processing, a speech model creates synthesized speech that corresponds to input data. The speech model may include an encoder for encoding the input data into a context vector and a decoder for decoding the context vector into spectrogram data. The speech model may further include a voice decoder that receives vocal characteristic data representing a desired vocal characteristic of synthesized speech. The voice decoder may process the vocal characteristic data to determine configuration data, such as weights, for use by the speech decoder.Type: GrantFiled: September 27, 2019Date of Patent: June 28, 2022Assignee: Amazon Technologies, Inc.Inventors: Roberto Barra Chicote, Vatsal Aggarwal, Andrew Paul Breen, Javier Gonzalez Hernandez, Nishant Prateek
-
Patent number: 11367431Abstract: A speech-processing system receives both text data and natural-understanding data (e.g., a domain, intent, and/or entity) related to a command represented in the text data. The system uses the natural-understanding data to vary vocal characteristics in determining spectrogram data corresponding to the text data based on the natural-understanding data.Type: GrantFiled: March 13, 2020Date of Patent: June 21, 2022Assignee: Amazon Technologies, Inc.Inventors: Antonio Bonafonte, Panagiotis Agis Oikonomou Filandras, Bartosz Perz, Arent van Korlaar, Ioannis Douratsos, Jonas Felix Ananda Rohnke, Elena Sokolova, Andrew Paul Breen, Nikhil Sharma
-
Publication number: 20210287656Abstract: A speech-processing system receives both text data and natural-understanding data (e.g., a domain, intent, and/or entity) related to a command represented in the text data. The system uses the natural-understanding data to vary vocal characteristics in determining spectrogram data corresponding to the text data based on the natural-understanding data.Type: ApplicationFiled: March 13, 2020Publication date: September 16, 2021Inventors: Antonio Bonafonte, Panagiotis Agis Oikonomou Filandras, Bartosz Perz, Arent van Korlaar, Ioannis Douratsos, Jonas Felix Ananda Rohnke, Elena Sokolova, Andrew Paul Breen, Nikhil Sharma
-
Patent number: 11017763Abstract: During text-to-speech processing, a sequence-to-sequence neural network model may process text data and determine corresponding spectrogram data. A normalizing flow component may then process this spectrogram data to predict corresponding phase data. An inverse Fourier transform may then be performed on the spectrogram and phase data to create an audio waveform that includes speech corresponding to the text.Type: GrantFiled: December 12, 2019Date of Patent: May 25, 2021Assignee: Amazon Technologies, Inc.Inventors: Vatsal Aggarwal, Nishant Prateek, Roberto Barra Chicote, Andrew Paul Breen
-
Publication number: 20210097976Abstract: During text-to-speech processing, a speech model creates synthesized speech that corresponds to input data. The speech model may include an encoder for encoding the input data into a context vector and a decoder for decoding the context vector into spectrogram data. The speech model may further include a voice decoder that receives vocal characteristic data representing a desired vocal characteristic of synthesized speech. The voice decoder may process the vocal characteristic data to determine configuration data, such as weights, for use by the speech decoder.Type: ApplicationFiled: September 27, 2019Publication date: April 1, 2021Inventors: Roberto Barra Chicote, Vatsal Aggarwal, Andrew Paul Breen, Javier Gonzalez Hernandez, Nishant Prateek
-
Publication number: 20200410981Abstract: A speech model is trained using multi-task learning. A first task may correspond to how well predicted audio matches training audio; a second task may correspond to a metric of perceived audio quality. The speech model may include, during training, layers related to the second task that are discarded at runtime.Type: ApplicationFiled: May 19, 2020Publication date: December 31, 2020Inventors: Thomas Edward Merritt, Adam Franciszek Nadolski, Nishant Prateek, Bartosz Putrycz, Roberto Barra Chicote, Vatsal Aggarwal, Andrew Paul Breen
-
Publication number: 20200394997Abstract: During text-to-speech processing, a speech model creates output audio data, including speech, that corresponds to input text data that includes a representation of the speech. A spectrogram estimator estimates a frequency spectrogram of the speech; the corresponding frequency-spectrogram data is used to condition the speech model. A plurality of acoustic features corresponding to different segments of the input text data, such as phonemes, syllable-level features, and/or word-level features, may be separately encoded into context vectors; the spectrogram estimator uses these separate context vectors to create the frequency spectrogram.Type: ApplicationFiled: July 7, 2020Publication date: December 17, 2020Inventors: Jaime Lorenzo Trueba, Thomas Renaud Drugman, Viacheslav Klimkov, Srikanth Ronanki, Thomas Edward Merritt, Andrew Paul Breen, Roberto Barra-Chicote
-
Publication number: 20200365137Abstract: A speech model includes a sub-model corresponding to a vocal attribute. The speech model generates an output waveform using a sample model, which receives text data, and a conditioning model, which receives text metadata and produces a prosody output for use by the sample model. If, during training or runtime, a different vocal attribute is desired or needed, the sub-model is re-trained or switched to a different sub-model corresponding to the different vocal attribute.Type: ApplicationFiled: June 23, 2020Publication date: November 19, 2020Inventors: Roberto Barra Chicote, Adam Franciszek Nadolski, Thomas Edward Merritt, Bartosz Putrycz, Andrew Paul Breen
-
Patent number: 10741169Abstract: During text-to-speech processing, a speech model creates output audio data, including speech, that corresponds to input text data that includes a representation of the speech. A spectrogram estimator estimates a frequency spectrogram of the speech; the corresponding frequency-spectrogram data is used to condition the speech model. A plurality of acoustic features corresponding to different segments of the input text data, such as phonemes, syllable-level features, and/or word-level features, may be separately encoded into context vectors; the spectrogram estimator uses these separate context vectors to create the frequency spectrogram.Type: GrantFiled: September 25, 2018Date of Patent: August 11, 2020Assignee: Amazon Technologies, Inc.Inventors: Jaime Lorenzo Trueba, Thomas Renaud Drugman, Viacheslav Klimkov, Srikanth Ronanki, Thomas Edward Merritt, Andrew Paul Breen, Roberto Barra-Chicote
-
Patent number: 10706837Abstract: A speech model includes a sub-model corresponding to a vocal attribute. The speech model generates an output waveform using a sample model, which receives text data, and a conditioning model, which receives text metadata and produces a prosody output for use by the sample model. If, during training or runtime, a different vocal attribute is desired or needed, the sub-model is re-trained or switched to a different sub-model corresponding to the different vocal attribute.Type: GrantFiled: June 13, 2018Date of Patent: July 7, 2020Assignee: Amazon Technologies, Inc.Inventors: Roberto Barra Chicote, Adam Franciszek Nadolski, Thomas Edward Merritt, Bartosz Putrycz, Andrew Paul Breen