AudioClassification

Classification model using transfer learning (Xception)

This is a dataset from an inclass kaggle competition. It contain various audio recordings and our task is to classify them into various sentiments. My approach was convering these audio files into images using Mel-frequency cepstral coefficients (MFCCs). Once converted into images we can use any image processing algorithms to classify them.

Data_Link :- https://drive.google.com/file/d/1dIj9s-oKFaUvHWujhwNOIRMCgYhSea7e/view?usp=sharing

File descriptions

1. Train - the training set

2. Test - the test set

3. submission.csv - a sample submission file in the correct format

4. map.csv - maps data to the labels

I don't have the labels for the test set but I still included the test files in the data.