Tornar als detalls de l'article Learning audio and image representations with bio-inspired trainable feature extractors Descarrega el PDF