Patents Assigned to Deepgram, Inc.
-
Patent number: 12380880Abstract: An end-to-end automatic speech recognition (ASR) system can be constructed by fusing a first ASR model with a transformer. The input of the transformer is a learned layer generated by the first ASR model. The fused ASR model and transformer can be treated as a single end-to-end model and trained as a single model. In some embodiments, the end-to-end speech recognition system can be trained using a teacher-student training technique by selectively truncating portions of the first ASR model and/or the transformer components and selectively freezing various layers during the training passes.Type: GrantFiled: April 3, 2023Date of Patent: August 5, 2025Assignee: Deepgram, Inc.Inventors: Andrew Nathan Seagraves, Deepak Subburam, Adam Joseph Sypniewski, Scott Ivan Stephenson, Jacob Edward Cutter, Michael Joseph Sypniewski, Daniel Lewis Shafer
-
Patent number: 12334075Abstract: Modern automatic speech recognition (ASR) systems can utilize artificial intelligence (AI) models to service ASR requests. The number and scale of AI models used in a modern ASR system can be substantial. The process of configuring and reconfiguring hardware to execute various AI models corresponding to a substantial number of ASR requests can be time consuming and inefficient. Among other features, the described technology utilizes batching of ASR requests, splitting of the ASR requests, and/or parallel processing to efficiently use hardware tasked with executing AI models corresponding to ASR requests. In one embodiment, the compute graphs of ASR tasks are used to batch the ASR requests. The corresponding AI models of each batch can be loaded into hardware, and batches can be processed in parallel. In some embodiments, the ASR requests are split, batched, and processed in parallel.Type: GrantFiled: October 14, 2022Date of Patent: June 17, 2025Assignee: Deepgram, Inc.Inventors: Adam Joseph Sypniewski, Joshua Gevirtz, Nikola Lazar Whallon, Anthony John Deschamps, Scott Ivan Stephenson
-
Patent number: 11676579Abstract: Systems and methods are disclosed for generating internal state representations of a neural network during processing and using the internal state representations for classification or search. In some embodiments, the internal state representations are generated from the output activation functions of a subset of nodes of the neural network. The internal state representations may be used for classification by training a classification model using internal state representations and corresponding classifications. The internal state representations may be used for search, by producing a search feature from an search input and comparing the search feature with one or more feature representations to find the feature representation with the highest degree of similarity.Type: GrantFiled: October 16, 2020Date of Patent: June 13, 2023Assignee: Deepgram, Inc.Inventors: Jeff Ward, Adam Sypniewski, Scott Stephenson
-
Patent number: 11367433Abstract: Systems and methods are disclosed for end-to-end neural networks for speech recognition and classification and additional machine learning techniques that may be used in conjunction or separately. Some embodiments comprise multiple neural networks, directly connected to each other to form an end-to-end neural network. One embodiment comprises a convolutional network, a first fully-connected network, a recurrent network, a second fully-connected network, and an output network. Some embodiments are related to generating speech transcriptions, and some embodiments relate to classifying speech into a number of classifications.Type: GrantFiled: May 29, 2020Date of Patent: June 21, 2022Assignee: Deepgram, Inc.Inventors: Adam Sypniewski, Jeff Ward, Scott Stephenson
-
Patent number: 10847138Abstract: Systems and methods are disclosed for generating internal state representations of a neural network during processing and using the internal state representations for classification or search. In some embodiments, the internal state representations are generated from the output activation functions of a subset of nodes of the neural network. The internal state representations may be used for classification by training a classification model using internal state representations and corresponding classifications. The internal state representations may be used for search, by producing a search feature from an search input and comparing the search feature with one or more feature representations to find the feature representation with the highest degree of similarity.Type: GrantFiled: May 21, 2019Date of Patent: November 24, 2020Assignee: Deepgram, Inc.Inventors: Jeff Ward, Adam Sypniewski, Scott Stephenson
-
Patent number: 10720151Abstract: Systems and methods are disclosed for end-to-end neural networks for speech recognition and classification and additional machine learning techniques that may be used in conjunction or separately. Some embodiments comprise multiple neural networks, directly connected to each other to form an end-to-end neural network. One embodiment comprises a convolutional network, a first fully-connected network, a recurrent network, a second fully-connected network, and an output network. Some embodiments are related to generating speech transcriptions, and some embodiments relate to classifying speech into a number of classifications.Type: GrantFiled: August 22, 2018Date of Patent: July 21, 2020Assignee: Deepgram, Inc.Inventors: Adam Sypniewski, Jeff Ward, Scott Stephenson
-
Publication number: 20200035222Abstract: Systems and methods are disclosed for end-to-end neural networks for speech recognition and classification and additional machine learning techniques that may be used in conjunction or separately. Some embodiments comprise multiple neural networks, directly connected to each other to form an end-to-end neural network. One embodiment comprises a convolutional network, a first fully-connected network, a recurrent network, a second fully-connected network, and an output network. Some embodiments are related to generating speech transcriptions, and some embodiments relate to classifying speech into a number of classifications.Type: ApplicationFiled: August 22, 2018Publication date: January 30, 2020Applicant: Deepgram, Inc.Inventors: Adam Sypniewski, Jeff Ward, Scott Stephenson
-
Publication number: 20200035224Abstract: Systems and methods are disclosed for generating internal state representations of a neural network during processing and using the internal state representations for classification or search. In some embodiments, the internal state representations are generated from the output activation functions of a subset of nodes of the neural network. The internal state representations may be used for classification by training a classification model using internal state representations and corresponding classifications. The internal state representations may be used for search, by producing a search feature from an search input and comparing the search feature with one or more feature representations to find the feature representation with the highest degree of similarity.Type: ApplicationFiled: May 21, 2019Publication date: January 30, 2020Applicant: Deepgram, Inc.Inventors: Jeff Ward, Adam Sypniewski, Scott Stephenson
-
Patent number: 10540959Abstract: Systems and methods are disclosed for customizing a neural network for a custom dataset, when the neural network has been trained on data from a general dataset. The neural network may comprise an output layer including one or more nodes corresponding to candidate outputs. The values of the nodes in the output layer may correspond to a probability that the candidate output is the correct output for an input. The values of the nodes in the output layer may be adjusted for higher performance when the neural network is used to process data from a custom dataset.Type: GrantFiled: December 26, 2018Date of Patent: January 21, 2020Assignee: Deepgram, Inc.Inventors: Jeff Ward, Adam Sypniewski, Scott Stephenson
-
Patent number: 10380997Abstract: Systems and methods are disclosed for generating internal state representations of a neural network during processing and using the internal state representations for classification or search. In some embodiments, the internal state representations are generated from the output activation functions of a subset of nodes of the neural network. The internal state representations may be used for classification by training a classification model using internal state representations and corresponding classifications. The internal state representations may be used for search, by producing a search feature from an search input and comparing the search feature with one or more feature representations to find the feature representation with the highest degree of similarity.Type: GrantFiled: August 22, 2018Date of Patent: August 13, 2019Assignee: Deepgram, Inc.Inventors: Jeff Ward, Adam Sypniewski, Scott Stephenson
-
Patent number: 10210860Abstract: Systems and methods are disclosed for customizing a neural network for a custom dataset, when the neural network has been trained on data from a general dataset. The neural network may comprise an output layer including one or more nodes corresponding to candidate outputs. The values of the nodes in the output layer may correspond to a probability that the candidate output is the correct output for an input. The values of the nodes in the output layer may be adjusted for higher performance when the neural network is used to process data from a custom dataset.Type: GrantFiled: August 22, 2018Date of Patent: February 19, 2019Assignee: Deepgram, Inc.Inventors: Jeff Ward, Adam Sypniewski, Scott Stephenson