Code available in Untitled3.ipynb
Audio-Classification model. Primary hurdle was preprocessing the audio signals to make them suitable for convolutional neural networks. Converted to mel spectograms which maps the signals from the amplitude-frequency domain to the frequency-time domain on a well popular audio processing scale called the mel scale. The model consists of convolutional layers followed by a few linear layers. Libraries used : 1)Pytorch 2)TorchAudio