P4-06: Singer Identity Representation Learning Using Self-Supervised Techniques
Bernardo Torres (Telecom Paris, Institut polytechnique de Paris)*, Stefan Lattner (Sony CSL), Gaël Richard (Telecom Paris, Institut polytechnique de Paris)
Subjects (starting with primary): MIR tasks -> music synthesis and transformation ; Knowledge-driven approaches to MIR -> representations of music ; Knowledge-driven approaches to MIR -> machine learning/artificial intelligence for music ; Musical features and properties -> timbre, instrumentation, and singing voice ; MIR tasks -> similarity metrics ; MIR tasks -> indexing and querying
Presented In Person: 4-minute short-format presentation
Significant strides have been made in creating voice identity representations using speech data. However, the same level of progress has not been achieved for singing voices. To bridge this gap, we suggest a framework for training singer identity encoders to extract representations suitable for various singing-related tasks, such as singing voice similarity and synthesis. We explore different self-supervised learning techniques on a large collection of isolated vocal tracks and apply data augmentations during training to ensure that the representations are invariant to pitch and content variations. We evaluate the quality of the resulting representations on singer similarity and identification tasks across multiple datasets, with a particular emphasis on out-of-domain generalization. Our proposed framework produces high-quality embeddings that outperform both speaker verification and wav2vec 2.0 pre-trained baselines on singing voice while operating at 44.1 kHz. We release our code and trained models to facilitate further research on singing voice and related areas.
If the video does not load properly please use the direct link to video