SyncTrack: Rhythmic Stability and Synchronization in Multi-Track Music Generation

Researchers introduce SyncTrack, an AI model for multi-track music generation that addresses rhythmic stability and synchronization issues in existing models. The model uses track-shared modules for common rhythm and track-specific modules for diverse timbres, introducing new metrics to evaluate multi-track music quality.

• →SyncTrack is a new AI model specifically designed for synchronous multi-track waveform music generation.
• →The model features track-shared modules for rhythm consistency and track-specific modules for instrument diversity.
• →Three new evaluation metrics were introduced: Inner-track Rhythmic Stability, Cross-track Beat Synchronization, and Cross-track Beat Dispersion.
• →Experiments show SyncTrack significantly improves multi-track music quality through enhanced rhythmic consistency.
• →The research addresses limitations in existing multi-track music generation models that focus on track differences rather than inherent properties.