TensorFlow Speech Recognition Kaggle Challenge to recognize simple spoken speech commands
| Authors: Divya Rajendran, Goutham Arra
Python3 is needed to run this code efficiently
pip install jupyter pandas numpy keras librosa sklearn tensorflow scipy tqdm pydub
Code for the project is available here
File: Report.pdf
-
We have utilized TensorFlow Speech Commands Data Sets [1] containing 65000, 30 second long speech signals as a training sample.
-
We have utilized 150,000 samples of test data, available in the Kaggle Challenge to test our model accuracy.
We have used LSTM + CNN model
TBD
TBD
[1]: "Warden P. Speech Commands: A public dataset for single-word speech recognition, 2017. Available from http://download.tensorflow.org/data/speech_commands_v0.01.tar.gz".