公开(公告)号：US12002454B2

公开(公告)日：2024-06-04

申请号：US17126703

申请日：2020-12-18

申请人： SWAMPFOX TECHNOLOGIES, INC.

发明人： Sergey A. Razin ,  Robert S. Cooper ,  Rick Ulmer ,  Tom Hanson

IPC分类号： G10L15/00 ,  G10L15/06 ,  G10L15/18 ,  G10L15/22 ,  G10L15/30 ,  G10L15/02 ,  G10L15/183

CPC分类号： G10L15/18 ,  G10L15/063 ,  G10L15/22 ,  G10L15/30 ,  G10L15/02 ,  G10L15/1822 ,  G10L15/183 ,  G10L2015/225

摘要： Embodiments of the innovation relate to, in a contact center apparatus, a method for recognizing user intent associated with user interaction with the contact center apparatus. The method includes receiving user interaction data; performing a feature extraction process on the user interaction data to generate feature data; performing an intent extraction operation on the feature data to extract topics included with the feature data; executing a classification engine on the topics extracted from the feature data to classify a user intent associated with the user interaction data; and directing the user to a corresponding working agent based upon the classified user intent

公开(公告)号：US20240177706A1

公开(公告)日：2024-05-30

申请号：US18515212

申请日：2023-11-20

申请人： Google LLC

发明人： Anshuman Tripathi ,  Soheil Khorram ,  Hasim Sak ,  Han Lu ,  Jaeyoung Kim ,  Qian Zhang

IPC分类号： G10L15/06 ,  G10L15/065 ,  G10L15/10

CPC分类号： G10L15/063 ,  G10L15/065 ,  G10L15/10 ,  G10L2015/0635

摘要： A method for training a sequence transduction model includes receiving a sequence of unlabeled input features extracted from unlabeled input samples. Using a teacher branch of an unsupervised subnetwork, the method includes processing the sequence of input features to predict probability distributions over possible teacher branch output labels, sampling one or more sequences of teacher branch output labels, and determining a sequence of pseudo output labels based on the one or more sequences of teacher branch output labels. Using a student branch that includes a student encoder of the unsupervised subnetwork, the method includes processing the sequence of input 10 features to predict probability distributions over possible student branch output labels, determining a negative log likelihood term based on the predicted probability distributions over possible student branch output labels and the sequence of pseudo output labels, and updating parameters of the student encoder.

公开(公告)号：US11996081B2

公开(公告)日：2024-05-28

申请号：US18324234

申请日：2023-05-26

申请人： Amazon Technologies, Inc.

发明人： Vasiliy Radostev ,  Ruhi Sarikaya ,  Rekha Seshadrinathan ,  Abhinav Sethy ,  Chetan Nagaraj Naik ,  Anjishnu Kumar

IPC分类号： G10L13/08 ,  G10L15/06 ,  G10L15/08 ,  G10L15/183

CPC分类号： G10L13/08 ,  G10L15/063 ,  G10L15/083 ,  G10L15/183

摘要： Techniques for generating a visual response to a user input are described. A system may receive a natural language input and use a machine learning model to determine a first component is to determine a response to the natural language input while a second component is to determine supplemental content related to the natural language input. The system may receive, from the first component, first image data corresponding to the response. The system may also receive, from the second component, second image data corresponding to the supplemental content. The system may send, to a display, a command to present the first image data and the second image data.

公开(公告)号：US11995404B2

公开(公告)日：2024-05-28

申请号：US16928519

申请日：2020-07-14

申请人： Microsoft Technology Licensing, LLC

发明人： Howard D'Souza ,  Regina Spitznagel ,  Debjani Sarkar

IPC分类号： G06F17/00 ,  G06F40/10 ,  G06F40/12 ,  G06F40/30 ,  G06Q10/10 ,  G16H15/00 ,  G16H40/20 ,  G10L15/06 ,  G16H10/60

CPC分类号： G06F40/30 ,  G06F40/10 ,  G06F40/12 ,  G06Q10/10 ,  G16H40/20 ,  G10L15/063 ,  G16H10/60 ,  G16H15/00

摘要： Techniques for training a natural language understanding (NLU) engine may include generating a first annotation of free-form text documenting a healthcare patient encounter and a link between the first annotation and a corresponding portion of the text, using the NLU engine. A second annotation of the text and a link between the second annotation and a corresponding portion of the text may be received from a human user. The first annotation and its corresponding link may be merged with the second annotation and its corresponding link. Training data may be provided to the engine in the form of the text and the merged annotations and links.

公开(公告)号：US11989514B2

公开(公告)日：2024-05-21

申请号：US17698018

申请日：2022-03-18

申请人： Capital One Services, LLC

发明人： Aysu Ezen Can ,  Zachary S. Brown ,  Chris Symons

IPC分类号： G06F40/30 ,  G06F40/166 ,  G06F40/284 ,  G10L15/06 ,  G10L15/16 ,  G10L15/22 ,  H04M3/42 ,  H04M3/51

CPC分类号： G06F40/284 ,  G06F40/166 ,  G10L15/063 ,  G10L15/16 ,  G10L15/22 ,  H04M3/42221 ,  H04M3/5175 ,  G06F40/30 ,  H04M2201/40

摘要： Disclosed herein are system, method, and computer program product embodiments for machine learning systems to process incoming call-center calls to provide communication summaries that capture effort levels of statements made during interactive communications. For a given call, the system receives a transcript as the input and generates a textual summary as the output. In order to improve a call summary and customize a summarization task to a call center domain, the technology disclosed herein may employ a classifier that predicts an effort level and attention score for individual utterances within a call transcript, ranks the attention scores and uses selected ones of the ranked utterances in the summary.

公开(公告)号：US20240161734A1

公开(公告)日：2024-05-16

申请号：US18282112

申请日：2022-03-31

申请人： COMMUNAUTE WOOPEN INC.

发明人： Edwin GRAPPIN ,  Jerome VERDIER

IPC分类号： G10L15/06

CPC分类号： G10L15/063

摘要： Method and servers for generating a speech model for generating signals representative of utterances in a first language based on signals representative of utterances in a second language are disclosed. The method comprises transmitting a first and a second speech models to a first and a second devices of a first and a second users respectively. The first device is communicatively coupled with the second device by an encrypted communication link. A third speech model is acquired from the second device based on a local training of the second speech model on the second device. A training set comprises a first and a second decrypted signals representative of an utterance of the first user in the first language and a translated utterance of the first user in the second language respectively. The speech model is locally generated by the server by combining the second and third speech models.

公开(公告)号：US20240161733A1

公开(公告)日：2024-05-16

申请号：US18054670

申请日：2022-11-11

申请人： International Business Machines Corporation

发明人： Ngoc Minh Tran ,  Hessel Tuinhof ,  Beat Buesser

IPC分类号： G10L15/06 ,  G10L15/01 ,  G10L15/065 ,  G10L25/18

CPC分类号： G10L15/063 ,  G10L15/01 ,  G10L15/065 ,  G10L25/18

摘要： A method, computer program product, and computer system for generation of training examples for training an automatic speech recognizer. Embodiments of the present invention can receive a training dataset of original audio signals and generate training examples for training an automatic speech recognizer based, at least in part, on a constructed imperceptible space for an original audio signal of the original audio signals and adversarial audio examples in the constructed imperceptible space. Embodiments of the present invention can then generate an imperceptible and adversarial audio example to an adversarial trainer for the automatic speech recognizer.

公开(公告)号：US11983551B2

公开(公告)日：2024-05-14

申请号：US18207053

申请日：2023-06-07

申请人： Apple Inc.

发明人： Taylor G. Carrigan ,  Patrick L. Coffman ,  David C. Graham

IPC分类号： G06F9/451 ,  G06F3/0481 ,  G06F3/0484 ,  G10L15/06 ,  G10L15/07 ,  G10L15/22 ,  G10L17/00

CPC分类号： G06F9/451 ,  G06F3/0481 ,  G06F3/0484 ,  G10L15/063 ,  G10L15/07 ,  G10L15/22 ,  G10L17/00 ,  G10L2015/0638 ,  G10L2015/223 ,  G10L2015/225

摘要： Examples of multi-user configuration are disclosed. An example method includes, at an electronic device: receiving a request; and in response to the request: if the voice input does not match a voice profile associated with an account associated with the electronic device: causing output of first information based on the request using a first account associated with the electronic device; if a setting of the electronic device has a first state, causing update of account data of the first account based on the request; and if the setting has a second state, forgoing causing update of the account data; and if the voice input matches a voice profile associated with an account associated with the electronic device: causing output of the first information using the account associated with the matching voice profile; and causing update of account data of the account based on the request.

公开(公告)号：US20240153396A1

公开(公告)日：2024-05-09

申请号：US18504880

申请日：2023-11-08

申请人： THE NONSENSE COMPANY, INC.

发明人： TYLER SLATER ,  JORDAN ELLETT ,  TRAVIS NUTTALL ,  ARIANNA Soltys ,  NICK HATHAWAY

IPC分类号： G09B5/06 ,  G06F40/47 ,  G09B19/06 ,  G10L15/06 ,  G10L15/16 ,  G10L15/19 ,  G10L15/30

CPC分类号： G09B5/065 ,  G06F40/47 ,  G09B19/06 ,  G10L15/063 ,  G10L15/16 ,  G10L15/19 ,  G10L15/30

摘要： Apparatuses, methods, systems, and program products are disclosed for language learning. An apparatus includes a processor and a memory that stores code executable by the processor to present a multimedia clip comprising an audio stream that is presented in a first language, receive a transcription of the audio stream in a second language different from the first language, determine one or more linguistic constructions of the received transcription that are being learned, and present at least a portion of the received transcription during playback of the multimedia clip with one or more portions of the presented transcription highlighted to emphasize the one or more linguistic constructions that are being learned.

公开(公告)号：US11978440B2

公开(公告)日：2024-05-07

申请号：US18323625

申请日：2023-05-25

申请人： Amazon Technologies, Inc.

发明人： Deepak Yavagal ,  Ajith Prabhakara ,  John Gray

IPC分类号： G10L15/183 ,  G10L15/06 ,  G10L15/22 ,  G10L15/08

CPC分类号： G10L15/183 ,  G10L15/063 ,  G10L15/22 ,  G10L2015/088

摘要： Techniques for processing input data for a detected user are described. Received image data is processed to identify an indicated user. Based on the user a machine learning model is implemented. The machine learning model is then used to process input data for a user input. An action is performed using the resulting output data.