Skip to content

Latest commit

 

History

History
42 lines (34 loc) · 1.53 KB

README.md

File metadata and controls

42 lines (34 loc) · 1.53 KB

forthebadge

made-with-bash made-with-python made-with-VSCode Generic badge

Easy Kaldi

Description

A collection of scripts based on Kaldi for speech recognition, diarization & language modeling

Speech Recognition asr

    1. Data prep
    1. Lexicon generation
    1. Grammar generation (pocolm & srilm)
    1. Feature extraction
    1. HMM-GMM training
    1. Data augmentation (speed, volume, reverb, music, noise, babble)
    1. Embedding (i-vector, x-vector)
    1. DNN training
    1. RNNLM training
    1. Rescoring

Diarization diarization

    1. i-vector (LIUM)
    1. x-vector (Kaldi)

Installation

Dependencies

  • depends on: Kaldi & slgasr
  • Refer to respective projects for install info

Free Datasets

Pretrained models

License

GPL

Authors

(c) 2020 Sylvain Le Groux [email protected]