TEAdapter: Supply Vivid Guidance for Controllable Text-to-Music Generation
Published in IEEE ICME 2024, 2024
We propose TEAdapter, a lightweight plugin for diffusion models that achieves fine-grained controllable music generation by extracting chord, melody, and instrument features from teacher music. We also design a structure-function-based (Intro/Chorus/Outro) multi-adapter collaboration and inpainting mechanism to address structural coherence in long audio generation.
Recommended citation: Jialing Zou*, Jiahao Mei*, Xudong Nan, et al. "TEAdapter: Supply Vivid Guidance for Controllable Text-to-Music Generation." IEEE ICME, 2024.
Download Paper