Efficient and Fast Generative-Based Singing Voice Separation using a Latent Diffusion Model | Publicación