[CVPR 2025] Taming Multimodal Joint Training for High-Quality Video-to-Audio Synthesis - View it on GitHub