About the Project:

Problem : Speech Recognition (Keyword Spotting)

• Data

• Implementation: I had taken inspiration from Kaggle Dataset, and implemented in Tensorflow 2.0.

Preprocessing:

• The proprocessing part, is loading the data, labelling the data, and then mixing the background noise with them, and then train the model, apply cross-validation technqiue, dimensionality reduction, and obtain the results.

Models used:

• The model used are:

1. Neural Network (Convolution Neural Network)
2. Random Forest
3. KNN

Flow of the Project:

There are three notebooks:

• KNN-Project-Recognition.py contains the preprocessing part, and then applying the KNN Model, and performing cross validation and dimensionality reduction
• Random-Forest-Project-Recognition.py contains the preprocessing part, and then applying the Random Forest Model, and performing cross validation and dimensionality reduction
• the whole code for speech recognition (neural network and random forest).py contains the preprocessing part, and then applying the Tensorflow Model and Random Forest, and dimensionality reduction