d-speech ~master

speech processing toolkit for D


To use this package, run the following command in your project's root directory:

D-SPEECH

CircleCI codecov

WIP. speech processing (especially, speech recognition) toolkit for D

feature

  • minimal
  • efficient
  • faster

roadmap

  • dsp.feature module
  • dsp.learn.acoustic module
  • dsp.wfst module

structure

  • source/dsp/
  • feature: speech feature (FBANK, MFCC) extraction module
  • learn: machine learning module
    • acoustic: acoustic model module
    • hmm: hidden Markov model based sequence transducer
    • gmm: Gaussian mixture model based acoustic model module
    • neural: nueral network based acoustic model module
    • language: language model module
    • ngram: n-gram based language model module
    • neural: neural network based language model module
  • wfst: weighted finite state transducer module
    • container: sequence acceptor and transducer containers
    • algorithm: wfst related algorithms
  • bin/
  • dsp-feature: speech feature (FBANK, MFCC) extractor command
Authors:
  • karita
Dependencies:
numir, snck, d-ffmpeg-light, d-graphviz, ggplotd, lubeck
Versions:
0.0.1 2019-Apr-16
~master 2019-Apr-16
Show all 2 versions
Download Stats:
  • 0 downloads today

  • 0 downloads this week

  • 0 downloads this month

  • 3 downloads total

Score:
0.5
Short URL:
d-speech.dub.pm