公开(公告)号：US20240330603A1

公开(公告)日：2024-10-03

申请号：US18309330

申请日：2023-04-28

申请人： Salesforce, Inc.

发明人： Lifu Tu ,  Yingbo Zhou ,  Caiming Xiong ,  Jin Qu

IPC分类号： G06F40/47 ,  G06F40/284 ,  G06F40/30 ,  G06F40/51

CPC分类号： G06F40/47 ,  G06F40/284 ,  G06F40/30 ,  G06F40/51

摘要： Embodiments described herein provide a method of training a language model by tuning a prompt. The method comprises masking tokens of first and second conversational texts which have the same semantic meaning but in different languages (e.g., a translation). The masked texts are input to a language model with a prepended soft prompt. The language model generates respective predicted outputs. A loss objective is computed including a masked language model loss. The prompt is updated based on the computed loss objective via backpropagation while keeping the language model frozen.

公开(公告)号：US12106059B2

公开(公告)日：2024-10-01

申请号：US18517720

申请日：2023-11-22

申请人： UNLIKELY ARTIFICIAL INTELLIGENCE LIMITED

发明人： William Tunstall-Pedoe ,  Finlay Curran ,  Harry Roscoe ,  Robert Heywood

IPC分类号： G06F17/00 ,  G06F16/242 ,  G06F16/31 ,  G06F16/332 ,  G06F16/951 ,  G06F40/123 ,  G06F40/126 ,  G06F40/20 ,  G06F40/205 ,  G06F40/211 ,  G06F40/226 ,  G06F40/242 ,  G06F40/279 ,  G06F40/30 ,  G06F40/35 ,  G06F40/45 ,  G06F40/47 ,  G06F40/58 ,  G06N3/0442 ,  G06N3/0455 ,  G06N3/0499 ,  G06N3/08 ,  G06N5/02 ,  G06Q10/1053 ,  G06Q30/0251 ,  G06Q30/0601 ,  G10L15/16 ,  G10L15/18 ,  G10L15/22 ,  G10L15/26 ,  G10L25/63 ,  G16H10/60 ,  H04L51/02 ,  G06N3/091 ,  G10L15/08

CPC分类号： G06F40/35 ,  G06F16/243 ,  G06F16/322 ,  G06F16/3329 ,  G06F16/951 ,  G06F40/123 ,  G06F40/126 ,  G06F40/20 ,  G06F40/205 ,  G06F40/211 ,  G06F40/226 ,  G06F40/242 ,  G06F40/279 ,  G06F40/30 ,  G06F40/45 ,  G06F40/47 ,  G06F40/58 ,  G06N3/0442 ,  G06N3/0455 ,  G06N3/0499 ,  G06N3/08 ,  G06N5/02 ,  G06Q10/1053 ,  G06Q30/0255 ,  G06Q30/0631 ,  G10L15/16 ,  G10L15/1815 ,  G10L15/22 ,  G10L15/26 ,  G10L25/63 ,  G16H10/60 ,  H04L51/02 ,  G06N3/091 ,  G10L2015/088

摘要： A computer implemented method for the automated analysis or use of data is implemented by a voice assistant. The method comprises the steps of: (a) storing in a memory a structured, machine-readable representation of data that conforms to a machine-readable language (‘machine representation’); the machine representation including representations of user speech or text input to a human/machine interface; and (b) automatically processing the machine representations to analyse the user speech or text input.

公开(公告)号：US20240320446A1

公开(公告)日：2024-09-26

申请号：US18124263

申请日：2023-03-21

申请人： JPMorgan Chase Bank, N.A.

发明人： Tomas DE LA ROSA TURBIDES ,  Daniel BORRAJO ,  Manuela VELOSO

IPC分类号： G06F40/47 ,  G06F40/169 ,  G06F40/186 ,  G06F40/205

CPC分类号： G06F40/47 ,  G06F40/169 ,  G06F40/186 ,  G06F40/205

摘要： A method for translating natural language data into constraints via memory-based processing is disclosed. The method includes receiving inputs via a graphical user interface, each of the inputs including input wording in a natural language format; parsing, by using a model, the inputs to retrieve a case from a case repository, the retrieval including identification of the case based on a similarity value and a predetermined similarity threshold; automatically adapting, by using the model, the retrieved case to the inputs; generating, based on a result of the adapting, constraints that characterize the inputs, the constraints relating to a rule that is mandated by the input; and evaluating the constraints.

公开(公告)号：US20240281617A1

公开(公告)日：2024-08-22

申请号：US18649227

申请日：2024-04-29

申请人： UNLIKELY ARTIFICIAL INTELLIGENCE LIMITED

发明人： William TUNSTALL-PEDOE ,  Finlay CURRAN ,  Harry ROSCOE ,  Robert HEYWOOD

IPC分类号： G06F40/35 ,  G06F16/242 ,  G06F16/31 ,  G06F16/332 ,  G06F16/951 ,  G06F40/123 ,  G06F40/126 ,  G06F40/20 ,  G06F40/205 ,  G06F40/211 ,  G06F40/226 ,  G06F40/242 ,  G06F40/279 ,  G06F40/30 ,  G06F40/45 ,  G06F40/47 ,  G06F40/58 ,  G06N3/0442 ,  G06N3/0455 ,  G06N3/0499 ,  G06N3/08 ,  G06N3/091 ,  G06N5/02 ,  G06Q10/1053 ,  G06Q30/0251 ,  G06Q30/0601 ,  G10L15/08 ,  G10L15/16 ,  G10L15/18 ,  G10L15/22 ,  G10L15/26 ,  G10L25/63 ,  G16H10/60 ,  H04L51/02

CPC分类号： G06F40/35 ,  G06F16/243 ,  G06F16/322 ,  G06F16/3329 ,  G06F16/951 ,  G06F40/123 ,  G06F40/126 ,  G06F40/20 ,  G06F40/205 ,  G06F40/211 ,  G06F40/226 ,  G06F40/242 ,  G06F40/279 ,  G06F40/30 ,  G06F40/45 ,  G06F40/47 ,  G06F40/58 ,  G06N3/0442 ,  G06N3/0455 ,  G06N3/0499 ,  G06N3/08 ,  G06N5/02 ,  G06Q10/1053 ,  G06Q30/0255 ,  G06Q30/0257 ,  G06Q30/0631 ,  G10L15/16 ,  G10L15/1815 ,  G10L15/22 ,  G10L15/26 ,  G10L25/63 ,  G16H10/60 ,  H04L51/02 ,  G06N3/091 ,  G10L2015/088

摘要： A computer implemented method for the automated analysis or use of data is implemented by a voice assistant. The method comprises the steps of: (a) storing in a memory a structured, machine-readable representation of data that conforms to a machine-readable language (‘machine representation’); the machine representation including representations of user speech or text input to a human/machine interface; and (b) automatically processing the machine representations to analyse the user speech or text input.

公开(公告)号：US20240257808A1

公开(公告)日：2024-08-01

申请号：US18435024

申请日：2024-02-07

申请人： Amazon Technologies, Inc.

发明人： Robert John Mars

IPC分类号： G10L15/22 ,  G06F40/47 ,  G10L13/02 ,  G10L15/08 ,  G10L15/18 ,  G10L15/30

CPC分类号： G10L15/22 ,  G06F40/47 ,  G10L13/02 ,  G10L15/08 ,  G10L15/18 ,  G10L15/30 ,  G10L2015/088 ,  G10L2015/223

摘要： A speech-processing system may provide access to one or more virtual assistants via a voice-controlled device. A user may leverage a first virtual assistant to translate a natural language command from a first language into a second language, which the device can forward to a second virtual assistant for processing. The device may receive a command from a user and send input data representing the command to a first speech-processing system representing the first virtual assistant. The device may receive a response in the form of a first natural language output from the first speech-processing system along with an indication that the first natural language output should be directed to a second speech-processing system representing the second virtual assistant. For example, the command may be in the first language, and the first natural language output may be in the second language, which is understandable by the second speech-processing system.

公开(公告)号：US20240202465A1

公开(公告)日：2024-06-20

申请号：US18540694

申请日：2023-12-14

申请人： CANON KABUSHIKI KAISHA

发明人： YOSUKE IGARASHI

IPC分类号： G06F40/47 ,  G06T11/20 ,  G06T11/60 ,  G06V10/94 ,  G06V20/70 ,  G06V30/19 ,  G06V30/30

CPC分类号： G06F40/47 ,  G06T11/20 ,  G06T11/60 ,  G06V10/945 ,  G06V20/70 ,  G06V30/191 ,  G06V30/30 ,  G06T2200/24

摘要： An information processing apparatus according to the present disclosure acquires a translation text by executing a translation process on an input text; selects one or more consecutive characters in the translation text as a selection text, based on a user operation; assigns a label selected based on a user operation, to a character string corresponding to the selection text in the input text; and stores the input text to which the label has been assigned, in a storage unit.

公开(公告)号：US20240177730A1

公开(公告)日：2024-05-30

申请号：US17927234

申请日：2022-11-21

申请人： Compass Pathfinder Limited

发明人： Robert F. Dougherty ,  Demeter Sztanko ,  Gregory A. Ryslik ,  Alexis Harrington ,  Andrew Bettke

IPC分类号： G10L25/66 ,  G06F40/47 ,  G10L15/26 ,  G10L17/26 ,  G16H20/70

CPC分类号： G10L25/66 ,  G06F40/47 ,  G10L15/26 ,  G10L17/26 ,  G16H20/70

摘要： Approaches for transcribing, translating, reviewing, tagging, and providing analytics for therapist sessions are provided. A media file that captures one or more interactions between one or more providers and a recipient of a service may be obtained. A transcript of at least a portion of the one or more interactions captured in the audio file may be generated. Using machine learning, a plurality of analytics may be inferred based, at least in part, upon content contained in the transcript. One or more biomarkers for the recipient may be determined based, at least in part, upon the plurality of analytics. A predicted response to the service to provide for display may be generated based, at least in part, upon the one or more biomarkers and the plurality of analytics.

公开(公告)号：US11995413B2

公开(公告)日：2024-05-28

申请号：US17070231

申请日：2020-10-14

申请人： LISUTO KK

发明人： Nir Platek ,  Pavel Zaslavsky

IPC分类号： G06F40/30 ,  G06F5/00 ,  G06F16/23 ,  G06F16/30 ,  G06F40/103 ,  G06F40/242 ,  G06F40/47 ,  G06F40/58 ,  G06Q10/087 ,  G06Q30/0601 ,  H04M3/487

CPC分类号： G06F40/47 ,  G06F5/00 ,  G06F16/235 ,  G06F16/30 ,  G06F40/242 ,  G06Q10/087 ,  G06Q30/0633 ,  H04M3/487

摘要： A method of transforming data including receiving data in a first language specific form, converting the data in the first language specific form to a language agnostic form, storing the data in the language agnostic form, converting the data in the language agnostic form to at least one second language specific form and exporting, on demand, the data in at least one of the at least one second language specific form.

公开(公告)号：US11989528B2

公开(公告)日：2024-05-21

申请号：US17553788

申请日：2021-12-16

申请人： YANDEX EUROPE AG

发明人： Anton Aleksandrovich Dvorkovich ,  Ivan Sergeevich Komarov

IPC分类号： G06F17/00 ,  G06F40/284 ,  G06F40/47 ,  G06N3/04

CPC分类号： G06F40/47 ,  G06F40/284 ,  G06N3/04

摘要： Methods and electronic devices for executing offline translation of a source word into a target word via a Neural Network an encoder and a decoder. The method includes splitting the source word into input tokens, generating vector representations for input tokens, and generating a first sequence of output tokens representative of a first candidate word. In response to the first candidate word not respecting at least one pre-determined rule, the method includes triggering the decoder to generate a second sequence of output tokens having a different at least one last output token than at least one last output token of the first sequence. The second sequence is representative of a second candidate word. In response to the second candidate word respecting the at least one pre-determined rule, the method includes determining that the second candidate word is the target word.

公开(公告)号：US20240153186A1

公开(公告)日：2024-05-09

申请号：US18411611

申请日：2024-01-12

申请人： Rovi Guides, Inc.

发明人： Yusuf AbdElhakam AbdElkader Marey ,  Reda Harb

IPC分类号： G06T13/40 ,  G06F40/47 ,  G06T13/20 ,  G06V40/16 ,  G06V40/20 ,  G09B21/00 ,  G10L15/18 ,  G10L15/22 ,  G10L21/10 ,  G10L25/63

CPC分类号： G06T13/40 ,  G06F40/47 ,  G06T13/205 ,  G06V40/174 ,  G06V40/20 ,  G09B21/009 ,  G10L15/1815 ,  G10L15/22 ,  G10L21/10 ,  G10L25/63 ,  G10L2021/065

摘要： Systems and methods for doing presenting an avatar that speaks sign language based on sentiment of a speaker is disclosed herein. A translation application running on a device receives a content item comprising a video and an audio, wherein the audio comprises a first plurality of spoken words in a first language. The video comprises a character speaking the first plurality of spoken words in the first language. The translation application translates the first plurality of spoken words of the first language into a first sign of a first sign language. The translation application determines an emotional state expressed by the character based on sentiment analysis. The translation application generates an avatar that speaks the first sign of the first sign language where the avatar exhibits the determined emotional state. The content item and the avatar are presented for display on the device.