Encodes a database of spoken digits to spectrogram images. Includes a keras implementation of a CNN trainer and predictor of the spoken digits dataset. Can conduct live predictions for testing by speaking digits into the mic
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
.gitignore
README.md
bgNoise.wav
config.py
datagen.py
dummy.wav
eval.py
getmean.py
googlenet.py
googlenet_custom_layers.py
main.py
model.py
module.py
predict.py
soundProcessing.py
test_processing.py
train.py
train_gen.py
wavToimg.py
wavVariations.py
webserver.py

README.md

Automatic Speech Recognition Training Platform

This tool helps in conducting experiments and evaluating performance of various ASR systems built using Deep Learning models.