Audio feature extraction and classification

Pytorch project accompanying the paper "Training Deep Pitch-Class Representations With a Multi-Label CTC Loss", ISMIR 2021.

Keras project accompanying the paper "Learning Pitch-Class Representations from Score--Audio Pairs of Classical Music", ISMIR 2021.

This Deep Learning project is develop to detect emergency vehicles (ambulance, police cars, ...) using street audio. CNN and chroma features will help us achieve this goal.

This project explores emotion recognition in audio data, focusing on feature extraction techniques while also comparing the performance of LSTM and 1D CNN models.