All Projects → funcwj → kaldi-python-io

funcwj / kaldi-python-io

Licence: Apache-2.0 license
A python IO interface for data accessing in kaldi

Programming Languages

python
139335 projects - #7 most used programming language

Projects that are alternatives of or similar to kaldi-python-io

Elpis
🙊 WIP software for creating speech recognition models.
Stars: ✭ 101 (+158.97%)
Mutual labels:  kaldi
Py Kaldi Asr
Some simple wrappers around kaldi-asr intended to make using kaldi's (online) decoders as convenient as possible.
Stars: ✭ 156 (+300%)
Mutual labels:  kaldi
Kaldi Active Grammar
Python Kaldi speech recognition with grammars that can be set active/inactive dynamically at decode-time
Stars: ✭ 196 (+402.56%)
Mutual labels:  kaldi
Kaldi Gop
Computes the GMM-based Goodness of Pronunciation (GOP). Bases on Kaldi.
Stars: ✭ 104 (+166.67%)
Mutual labels:  kaldi
Speech To Text Russian
Проект для распознавания речи на русском языке на основе pykaldi.
Stars: ✭ 151 (+287.18%)
Mutual labels:  kaldi
Kaldiio
A pure python module for reading and writing kaldi ark files
Stars: ✭ 160 (+310.26%)
Mutual labels:  kaldi
Factorized Tdnn
PyTorch implementation of the Factorized TDNN (TDNN-F) from "Semi-Orthogonal Low-Rank Matrix Factorization for Deep Neural Networks" and Kaldi
Stars: ✭ 98 (+151.28%)
Mutual labels:  kaldi
Zeroth
Kaldi-based Korean ASR (한국어 음성인식) open-source project
Stars: ✭ 248 (+535.9%)
Mutual labels:  kaldi
Eend
End-to-End Neural Diarization
Stars: ✭ 153 (+292.31%)
Mutual labels:  kaldi
Kaldi Onnx
Kaldi model converter to ONNX
Stars: ✭ 174 (+346.15%)
Mutual labels:  kaldi
Tf Kaldi Speaker
Neural speaker recognition/verification system based on Kaldi and Tensorflow
Stars: ✭ 117 (+200%)
Mutual labels:  kaldi
Kaldi
kaldi-asr/kaldi is the official location of the Kaldi project.
Stars: ✭ 11,151 (+28492.31%)
Mutual labels:  kaldi
Ctc pytorch
CTC end -to-end ASR for timit and 863 corpus.
Stars: ✭ 161 (+312.82%)
Mutual labels:  kaldi
Vosk Api
Offline speech recognition API for Android, iOS, Raspberry Pi and servers with Python, Java, C# and Node
Stars: ✭ 1,357 (+3379.49%)
Mutual labels:  kaldi
Setk
Tools for Speech Enhancement integrated with Kaldi
Stars: ✭ 227 (+482.05%)
Mutual labels:  kaldi
Pytorch Kaldi Neural Speaker Embeddings
A light weight neural speaker embeddings extraction based on Kaldi and PyTorch.
Stars: ✭ 99 (+153.85%)
Mutual labels:  kaldi
Pykaldi2
Yet another speech toolkit based on Kaldi and PyTorch
Stars: ✭ 158 (+305.13%)
Mutual labels:  kaldi
asr24
24-hour Automatic Speech Recognition
Stars: ✭ 27 (-30.77%)
Mutual labels:  kaldi
Lhotse
Stars: ✭ 236 (+505.13%)
Mutual labels:  kaldi
Pytorch Kaldi
pytorch-kaldi is a project for developing state-of-the-art DNN/RNN hybrid speech recognition systems. The DNN part is managed by pytorch, while feature extraction, label computation, and decoding are performed with the kaldi toolkit.
Stars: ✭ 2,097 (+5276.92%)
Mutual labels:  kaldi

Kaldi Python IO

A python (3.6+) wrapper for Kaldi's data accessing.

Support Type

  • Kaldi's binary archives (*.ark)
  • Kaldi's scripts (alignments & features, *.scp)
  • Kaldi nnet3 data examples in binary (*.egs)

Install

python setup.py install or pip install kaldi-python-io

Usage

  • ArchiveReader && AlignArchiveReader

    # allow only sequential index
    ark_reader = ArchiveReader("copy-feats ark:foo.ark ark:- |")
    for key, _ in ark_reader:
        print(key)
    ali_reader = AlignArchiveReader("gunzip -c foo.ali.gz |")
    for key, _ in ark_reader:
        print(key)
  • Nnet3EgsReader

    # allow only sequential index
    egs_reader = Nnet3EgsReader("foo.egs")
    for key, _ in egs_reader:
        print(key)
  • ArchiveWriter

    with ArchiveWriter("foo.ark", "foo.scp") as writer:
        for i in range(10):
            mat = np.random.rand(100, 20)
            writer.write(f"mat-{i}", mat)
  • ScriptReader && AlignScriptReader

    # allow sequential/random index
    scp_reader = ScriptReader("shuf foo.scp | head -n 2")
    for key, mat in scp_reader:
        print(f"{key}: {mat.shape}")
    ali_reader = AlignScriptReader("foo.ali.scp")
    for key, ali in ali_reader:
        print(f"{key}: {ali.shape}")
Note that the project description data, including the texts, logos, images, and/or trademarks, for each open source project belongs to its rightful owner. If you wish to add or remove any projects, please contact us at [email protected].