Deep Learning Tools for Speech Recognition

writting in Python and running on GPUs. This tool was developed with a focus on enabling fast experimentation for Speech Recognition. Being able to go from idea to result with the least possible delay is key to doing good research.

Moved to https://bitbucket.org/AustinZhang/deep-speech as a private repo, please email me for access.

Dependencies: Theano and keras

• Install Theano

sudo pip install git+git://github.com/Theano/Theano.git

• Install keras

sudo pip install keras

---

1. Speech_IO:

supports read and write Feature files, Label files, Model files and Lattices

speech_io/

feature_io.py

label_io.py

model_io.py

lattice_io.py

---

2. Model Examples:

DNN, CNN, LSTM, CLDNN, FFNN, Attention models

---

3. Recipes on ASR tasks:

Switchboard, TIMIT