Patent Leaderboard
USPTO Patent Rankings Data through Dec 31, 2025
Home › Method and system for joint training of hybrid neural networks for acoustic modeling in automatic speech recognition

Method and system for joint training of hybrid neural networks for acoustic modeling in automatic speech recognition

US Patent 9665823 · Granted May 30, 2017

Estimated economic value: $1,983,000 ⓘ

Assignee

Inventors

View full patent text on Google Patents →

More Patents by These Inventors

Length perturbation techniques for improving generalization of deep neural netwo2025
Textual knowledge transfer for improved speech recognition and understanding2025
Multi-speaker data augmentation for improved end-to-end automatic speech recogni2025
Accuracy of streaming RNN transducer2025
Reducing exposure bias in machine learning training of sequence-to-sequence tran2024
Integrating dialog history into end-to-end spoken language understanding systems2024
Training end-to-end spoken language understanding systems with unordered entitie2024
Joint automatic speech recognition and speaker diarization2024
Chunking and overlap decoding strategy for streaming RNN transducers for speech 2024
Integrating text inputs for training and adapting neural network transducer ASR 2024
Customization of recurrent neural network transducers for speech recognition2024
Accuracy of streaming RNN transducer2023