Listen carefully and tell: an audio captioning system based on residual learning and gammatone audio representation | Publicación