Feature Learning in Medical Acoustics Implementation of the comparision between learnable frontends (LEAF and nnAudio) and fixed frontend (Mel-filterbank).