Skip to content

zhaoyue-zephyrus/RecurrentConvNet-for-Speech

master
Switch branches/tags

Name already in use

A tag already exists with the provided branch name. Many Git commands accept both tag and branch names, so creating this branch may cause unexpected behavior. Are you sure you want to create this branch?
Code

Latest commit

 

Git stats

Files

Permalink
Failed to load latest commit information.
Type
Name
Latest commit message
Commit time
 
 

This repo holds the codes for the paper

Recurrent Convolutional Neural Network For Speech Processing, Yue Zhao, Xingyu Jin, and Xiaolin Hu, to appear in the 42nd IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2017), New Orleans, USA.

Documentation

Install kaldi

Clone Kaldi. Follow the documentation to build the toolkit. (Remember to install IRSTLM via ./tools/extras/install_irstlm.sh since the default setting does not include it.)

Install CNTK

Install CNTK.

Run experiments on timit

Put the scripts in timit-scripts/ under egs/timit/s5/ in the Kaldi repository.

Run ./run.sh to produce a HMM model by tri3 system.

Run ./train_rcnn_tri00.sh to generate the RCNN model described in our paper.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages