Music Foundation Model as Generic Booster for Music Downstream Tasks
This work addresses the challenge of data scarcity and performance limitations in various music downstream tasks for researchers and practitioners in music processing.
The authors tackled the problem of improving performance across diverse music processing tasks by using hierarchical intermediate features from a single music foundation model called SoniDo. Their approach achieved enhanced results on tasks including music tagging, transcription, source separation, and mixing, though no concrete numbers were provided.
We demonstrate the efficacy of using intermediate representations from a single foundation model to enhance various music downstream tasks. We introduce SoniDo, a music foundation model (MFM) designed to extract hierarchical features from target music samples. By leveraging hierarchical intermediate features, SoniDo constrains the information granularity, leading to improved performance across various downstream tasks including both understanding and generative tasks. We specifically evaluated this approach on representative tasks such as music tagging, music transcription, music source separation, and music mixing. Our results reveal that the features extracted from foundation models provide valuable enhancements in training downstream task models. This highlights the capability of using features extracted from music foundation models as a booster for downstream tasks. Our approach not only benefits existing task-specific models but also supports music downstream tasks constrained by data scarcity. This paves the way for more effective and accessible music processing solutions.