Remixing Music for Hearing Aids Using Ensemble of Fine-Tuned Source Separators
This work addresses music enhancement for hearing aid users, representing an incremental improvement through fine-tuning and ensemble methods.
The paper tackled the problem of remixing and enhancing music for hearing aid users, achieving first place in the Cadenza ICASSP 2024 Grand Challenge with the best average Hearing-Aid Audio Quality Index (HAAQI) score on the evaluation dataset.
This paper introduces our system submission for the Cadenza ICASSP 2024 Grand Challenge, which presents the problem of remixing and enhancing music for hearing aid users. Our system placed first in the challenge, achieving the best average Hearing-Aid Audio Quality Index (HAAQI) score on the evaluation data set. We describe the system, which uses an ensemble of deep learning music source separators that are fine tuned on the challenge data. We demonstrate the effectiveness of our system through the challenge results and analyze the importance of different system aspects through ablation studies.