Multi-Stage Music Source Restoration with BandSplit-RoFormer Separation and HiFi++ GAN
This work addresses the challenge of music source restoration for audio engineers and researchers, providing a system for the MSR ICASSP Challenge 2025.
This paper tackles the problem of music source restoration (MSR), aiming to recover original instrument stems from mixed and mastered audio. The proposed system decomposes MSR into separation and restoration, using a BandSplit-RoFormer for 8-stem separation and a HiFi++ GAN for waveform restoration.
Music Source Restoration (MSR) targets recovery of original, unprocessed instrument stems from fully mixed and mastered audio, where production effects and distribution artifacts violate common linear-mixture assumptions. This technical report presents the CP-JKU team's system for the MSR ICASSP Challenge 2025. Our approach decomposes MSR into separation and restoration. First, a single BandSplit-RoFormer separator predicts eight stems plus an auxiliary other stem, and is trained with a three-stage curriculum that progresses from 4-stem warm-start fine-tuning (with LoRA) to 8-stem extension via head expansion. Second, we apply a HiFi++ GAN waveform restorer trained as a generalist and then specialized into eight instrument-specific experts.