Unsupervised Source Separation By Steering Pretrained Music Models
We showcase an unsupervised method that repurposes deep models trained for music generation and music tagging for audio source separation, without any retraining. An audio generation model is conditioned on an input mixture, producing a latent encoding of the audio used to generate audio. This generated audio is fed to a pretrained music tagger that creates source labels. The cross-entropy loss between the tag distribution for the generated audio and a predefined distribution for an isolated source is used to guide gradient ascent in the (unchanging) latent space of the generative model. This system does not update the weights of the generative model or the tagger, and only relies on moving through the generative model's latent space to produce separated sources. We use OpenAI's Jukebox as the pretrained generative model, and we couple it with four kinds of pretrained music taggers (two architectures and two tagging datasets). Experimental results on two source separation datasets, show this approach can produce separation estimates for a wider variety of sources than any tested supervised or unsupervised system. This work points to the vast and heretofore untapped potential of large pretrained music models for audio-to-audio tasks like source separation.
Code (1)
Tasks
Audio GenerationAudio Source SeparationMusic GenerationMusic TaggingTAGMethods 이 논문이 사용한 방법론
Similar Papers 제목 키워드 기반
Unsupervised Music Source Separation Using Differentiable Parametric Source Models
Supervised deep learning approaches to underdetermined audio source separation achieve state-of-the-art performance but require a dataset of mixtures along with their corresponding isolated source signals. Such datasets …
Audio Source SeparationDeep LearningMusic Source SeparationVocal ensemble separationA Study of Transfer Learning in Music Source Separation
Supervised deep learning methods for performing audio source separation can be very effective in domains where there is a large amount of training data. While some music domains have enough data suitable for training a s…
Audio Source SeparationData AugmentationDeep LearningMusic Source Separation+1Pac-HuBERT: Self-Supervised Music Source Separation via Primitive Auditory Clustering and Hidden-Unit BERT
In spite of the progress in music source separation research, the small amount of publicly-available clean source data remains a constant limiting factor for performance. Thus, recent advances in self-supervised learning…
ClusteringDecoderMusic Source SeparationSelf-Supervised LearningIs MixIT Really Unsuitable for Correlated Sources? Exploring MixIT for Unsupervised Pre-training in Music Source Separation
In music source separation (MSS), obtaining isolated sources or stems is highly costly, making pre-training on unlabeled data a promising approach. Although source-agnostic unsupervised learning like mixture-invariant tr…
Music Source SeparationUnsupervised Pre-trainingUnsupervised Interpretable Representation Learning for Singing Voice Separation
In this work, we present a method for learning interpretable music signal representations directly from waveform signals. Our method can be trained using unsupervised objectives and relies on the denoising auto-encoder m…
DenoisingMusic Source SeparationRepresentation Learning