Audio foundation models for gunshot characterization: a self-supervised learning approach | Publicación