中科微点-全球专利检索

  1. Login
  2. Sign Up

Search Results

25 records (took 0.029 seconds)

    Clockwork Hierarchal Variational Encoder
    11.
    发明申请
    Clockwork Hierarchal Variational Encoder 有权

    公开(公告)号：US20220172705A1

    公开(公告)日：2022-06-02

    申请号：US17650452

    申请日：2022-02-09

    Applicant: Google LLC

    Inventor： Robert Andrew James Clark ,  Chun-an Chan ,  Vincent Ping Leung Wan

    IPC: G10L15/06 ,  G10L15/22 ,  G10L15/16 ,  G10L25/18 ,  G10L25/24 ,  G10L15/02 ,  G06N3/04 ,  G06N3/08 ,  G10L25/21

    Abstract: A method for providing a frame-based mel spectral representation of speech includes receiving a text utterance having at least one word, and selecting a mel spectral embedding for the text utterance. Each word in the text utterance has at least one syllable and each syllable has at least one phoneme. For each phoneme, using the selected mel spectral embedding, the method also includes: predicting a duration of the corresponding phoneme by encoding linguistic features of the corresponding phoneme with a corresponding syllable embedding for the syllable that includes the corresponding phoneme; and generating a plurality of fixed-length predicted mel-frequency spectrogram frames based on the predicted duration for the corresponding phoneme. Each fixed-length predicted mel-frequency spectrogram frame representing mel-spectral information of the corresponding phoneme.

    Speech Synthesis Prosody Using A BERT Model
    12.
    发明申请
    Speech Synthesis Prosody Using A BERT Model 有权

    公开(公告)号：US20210350795A1

    公开(公告)日：2021-11-11

    申请号：US16867427

    申请日：2020-05-05

    Applicant: Google LLC

    Inventor： Tom Marius Kenter ,  Manish Kumar Sharma ,  Robert Andrew James Clark ,  Aliaksei Severyn

    IPC: G10L15/16 ,  G10L15/02 ,  G06N3/08 ,  G10L15/06

    Abstract: A method for generating a prosodic representation includes receiving a text utterance having one or more words. Each word has at least one syllable having at least one phoneme. The method also includes generating, using a Bidirectional Encoder Representations from Transformers (BERT) model, a sequence of wordpiece embeddings and selecting an utterance embedding for the text utterance, the utterance embedding representing an intended prosody. Each wordpiece embedding is associated with one of the one or more words of the text utterance. For each syllable, using the selected utterance embedding and a prosody model that incorporates the BERT model, the method also includes generating a corresponding prosodic syllable embedding for the syllable based on the wordpiece embedding associated with the word that includes the syllable and predicting a duration of the syllable by encoding linguistic features of each phoneme of the syllable with the corresponding prosodic syllable embedding for the syllable.

    TRAINING NEURAL NETWORKS TO GENERATE STRUCTURED EMBEDDINGS
    14.
    发明申请
    TRAINING NEURAL NETWORKS TO GENERATE STRUCTURED EMBEDDINGS 有权

    公开(公告)号：US20230060886A1

    公开(公告)日：2023-03-02

    申请号：US18049995

    申请日：2022-10-26

    Applicant: Google LLC

    Inventor： Robert Andrew James Clark ,  Chun-an Chan ,  Vincent Ping Leung Wan

    IPC: G06N20/00 ,  G06N3/04 ,  G06N3/08

    Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training a machine learning model to generate embeddings of inputs to the machine learning model, the machine learning model having an encoder that generates the embeddings from the inputs and a decoder that generates outputs from the generated embeddings, wherein the embedding is partitioned into a sequence of embedding partitions that each includes one or more dimensions of the embedding, the operations comprising: for a first embedding partition in the sequence of embedding partitions: performing initial training to train the encoder and a decoder replica corresponding to the first embedding partition; for each particular embedding partition that is after the first embedding partition in the sequence of embedding partitions: performing incremental training to train the encoder and a decoder replica corresponding to the particular partition.

    Clockwork Hierarchical Variational Encoder
    17.
    发明申请
    Clockwork Hierarchical Variational Encoder 审中-公开

    公开(公告)号：US20200074985A1

    公开(公告)日：2020-03-05

    申请号：US16678981

    申请日：2019-11-08

    Applicant: Google LLC

    Inventor： Robert Andrew James Clark ,  Chun-an Chan ,  Vincent Ping Leung Wan

    IPC: G10L15/06 ,  G10L15/22 ,  G10L15/16 ,  G10L25/18 ,  G10L25/24 ,  G10L25/21 ,  G10L15/02 ,  G06N3/04 ,  G06N3/08

    Abstract: A method for providing a frame-based mel spectral representation of speech includes receiving a text utterance having at least one word, and selecting a mel spectral embedding for the text utterance. Each word in the text utterance has at least one syllable and each syllable has at least one phoneme. For each phoneme, using the selected mel spectral embedding, the method also includes: predicting a duration of the corresponding phoneme by encoding linguistic features of the corresponding phoneme with a corresponding syllable embedding for the syllable that includes the corresponding phoneme; and generating a plurality of fixed-length predicted mel-frequency spectrogram frames based on the predicted duration for the corresponding phoneme. Each fixed-length predicted mel-frequency spectrogram frame representing mel-spectral information of the corresponding phoneme.

    Two-level text-to-speech systems using synthetic training data
    19.
    发明授权
    Two-level text-to-speech systems using synthetic training data 有权

    公开(公告)号：US12260851B2

    公开(公告)日：2025-03-25

    申请号：US17305809

    申请日：2021-07-14

    Applicant: Google LLC

    Inventor： Lev Finkelstein ,  Chun-an Chan ,  Byungha Chun ,  Norman Casagrande ,  Yu Zhang ,  Robert Andrew James Clark ,  Vincent Wan

    IPC: G10L13/00 ,  G10L13/047 ,  G10L13/08

    Abstract: A method includes obtaining training data including a plurality of training audio signals and corresponding transcripts. Each training audio signal is spoken by a target speaker in a first accent/dialect. For each training audio signal of the training data, the method includes generating a training synthesized speech representation spoken by the target speaker in a second accent/dialect different than the first accent/dialect and training a text-to-speech (TTS) system based on the corresponding transcript and the training synthesized speech representation. The method also includes receiving an input text utterance to be synthesized into speech in the second accent/dialect. The method also includes obtaining conditioning inputs that include a speaker embedding and an accent/dialect identifier that identifies the second accent/dialect. The method also includes generating an output audio waveform corresponding to a synthesized speech representation of the input text sequence that clones the voice of the target speaker in the second accent/dialect.

Patent Agency Ranking