Artifacts and noise - Githubissues

andabi / music-source-separation

Deep neural networks for separating singing voice from music written in TensorFlow

795 stars 150 forks source link

Artifacts and noise #30

Open prathmat opened 5 years ago

prathmat commented 5 years ago

Hello, Can somebody help me regarding the reduction of artifacts and noise in the separated audio sources, Iam getting a lot of high frequency artifacts , is there a way to modify the cost function ?