Abstract: https://docs.google.com/document/d/1ZuANeFxBhUZMn2YhAssWtF9ALVaMNd1GFn1Toyv46TA/edit#heading=h.nj23sjpj5u97
The PS was on Audio classification using CNNs. Each audio file had a text file associated with it, which contained the intervals of each cycle of breathing along with its labels for wheezes and crackles.
So we iterated through the audio files and in each iteration, we sliceD the audio file into each respective cycles with respect to the time stamps given in the text file associated. Then we converted these sliced audio files into spectrogram images using Librosa library of Python.
