🎬🔊 ControlFoley — Unified & Controllable Video-to-Audio
Generate synchronized sound effects from video, with optional control over what the video should sound like through a text prompt or reference audio.
One unified model supports video-to-audio, text-guided video-to-audio, text-controlled video-to-audio, reference-audio-controlled video-to-audio, and text-to-audio.
Accepted at ACM MM 2026 Project repository
If useful, please ❤️ this Space and ⭐ the repository.
1 10
1 10
10 50
Examples
| Task type | Input video (optional) | Reference audio (optional) | Prompt (optional) |
|---|