Patents by Inventor Yuxiang KONG

Yuxiang KONG 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).

  • Method of training speech recognition model, electronic device and storage medium
    Patent number: 12260852
    Abstract: A method of training a speech recognition model is provided. The method includes that: speech data of each of a plurality of training samples is inputted into a teacher model and a to-be-trained speech recognition model separately. Additionally, an embedding outputted by the teacher model and encoded data outputted by the to-be-trained speech recognition model are obtained. Furthermore, quantized codebook data is obtained by performing a multi-codebook quantization on the embedding. A loss is calculated based on the encoded data, the quantized codebook data, and text data in the training sample. Moreover, a trained speech recognition model is obtained by stopping training the to-be-trained speech recognition model when the loss is less than or equal to a preset loss threshold and/or trained times is greater than preset trained times.
    Type: Grant
    Filed: December 9, 2022
    Date of Patent: March 25, 2025
    Assignee: BEIJING XIAOMI MOBILE SOFTWARE CO., LTD.
    Inventors: Zengwei Yao, Liyong Guo, Povey Daniel, Long Lin, Fangjun Kuang, Wei Kang, Mingshuang Luo, Quandong Wang, Yuxiang Kong
  • METHOD OF TRAINING SPEECH RECOGNITION MODEL, ELECTRONIC DEVICE AND STORAGE MEDIUM
    Publication number: 20230386448
    Abstract: A method of training a speech recognition model is provided. The method includes that: speech data of each of a plurality of training samples is inputted into a teacher model and a to-be-trained speech recognition model separately. Additionally, an embedding outputted by the teacher model and encoded data outputted by the to-be-trained speech recognition model are obtained. Furthermore, quantized codebook data is obtained by performing a multi-codebook quantization on the embedding. A loss is calculated based on the encoded data, the quantized codebook data, and text data in the training sample. Moreover, a trained speech recognition model is obtained by stopping training the to-be-trained speech recognition model when the loss is less than or equal to a preset loss threshold and/or trained times is greater than preset trained times.
    Type: Application
    Filed: December 9, 2022
    Publication date: November 30, 2023
    Applicant: BEIJING XIAOMI MOBILE SOFTWARE CO., LTD.
    Inventors: Zengwei YAO, Liyong GUO, POVEY DANIEL, Long LIN, Fangjun KUANG, Wei KANG, Mingshuang LUO, Quandong WANG, Yuxiang KONG