Weakly Supervised Audio Source Separation via Spectrum Energy Preserved Wasserstein Learning
| Autor: | Yu Chen Zhou, Ning Zhang, Junchi Yan |
|---|---|
| Rok vydání: | 2018 |
| Předmět: |
FOS: Computer and information sciences
Sound (cs.SD) Property (programming) Computer science business.industry Model selection 020206 networking & telecommunications Pattern recognition 02 engineering and technology Function (mathematics) Regularization (mathematics) Computer Science - Sound Term (time) 030507 speech-language pathology & audiology 03 medical and health sciences Audio and Speech Processing (eess.AS) FOS: Electrical engineering electronic engineering information engineering 0202 electrical engineering electronic engineering information engineering Source separation Benchmark (computing) Artificial intelligence 0305 other medical science business Energy (signal processing) Electrical Engineering and Systems Science - Audio and Speech Processing |
| Zdroj: | IJCAI |
| DOI: | 10.24963/ijcai.2018/636 |
| Popis: | Separating audio mixtures into individual instrument tracks has been a standing challenge. We introduce a novel weakly supervised audio source separation approach based on deep adversarial learning. Specifically, our loss function adopts the Wasserstein distance which directly measures the distribution distance between the separated sources and the real sources for each individual source. Moreover, a global regularization term is added to fulfill the spectrum energy preservation property regardless separation. Unlike state-of-the-art weakly supervised models which often involve deliberately devised constraints or careful model selection, our approach need little prior model specification on the data, and can be straightforwardly learned in an end-to-end fashion. We show that the proposed method performs competitively on public benchmark against state-of-the-art weakly supervised methods. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|