Research map / Video / Video generation
Foley Music: Learning to Generate Music from Videos
Chuang Gan, Deng Huang, Peihao Chen, Joshua B. Tenenbaum, Antonio Torralba
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use skeleton keypoints instead of RGB image features (ablation) | Foley Music · URMP (violin) | NLL of MIDI events: 1.586 -> 1.558 | Improved |
| use skeleton keypoints instead of RGB image features (ablation) | Foley Music · URMP (trumpet) | NLL of MIDI events: 2.748 -> 2.512 | Improved |
| use skeleton keypoints instead of optical flow (ablation) | Foley Music · URMP (trumpet) | NLL of MIDI events: 3.013 -> 2.512 | Improved |
| use skeleton keypoints instead of optical flow (ablation) | Foley Music · URMP (viola) | NLL of MIDI events: 3.859 -> 3.603 | Improved |
| use music transformer instead of GRU (ablation) | Foley Music · URMP (trumpet) | NLL of MIDI events: 2.631 -> 2.512 | Improved |
| use music transformer instead of GRU (ablation) | Foley Music · URMP (violin) | NLL of MIDI events: 1.631 -> 1.558 | Improved |
| add hand keypoints to body keypoints (ablation) | Foley Music · URMP (trumpet) | NLL of MIDI events: 2.805 -> 2.512 | Improved |
| add hand keypoints to body keypoints (ablation) | Foley Music · URMP (flute) | NLL of MIDI events: 2.259 -> 1.995 | Improved |
| add hand keypoints to body keypoints (ablation) | Foley Music · URMP (violin) | NLL of MIDI events: 1.565 -> 1.558 | Improved |
| transformer without hand keypoints versus GRU (ablation) | Foley Music · URMP (trumpet) | NLL of MIDI events: 2.631 -> 2.805 | Hurt |
| MIDI-based generation instead of GAN, SampleRNN, WaveNet | Foley Music · MUSIC/AtinPiano human real-vs-fake | success rate: 12% -> 38% | Improved |
| MIDI-based generation instead of GAN, SampleRNN, WaveNet | Foley Music · MUSIC/AtinPiano generated sound | NDB: 25 -> 20 | Improved |
Closest papers by meaning
- V2Meow: Meowing to the Visual Beat via Video-to-Music Generation (AAAI 2024 · similarity 0.88)
- MIDI-GPT: A Controllable Generative Model for Computer-Assisted Multitrack Music Composition (AAAI 2025 · similarity 0.88)
- VidMuse: A Simple Video-to-Music Generation Framework with Long-Short-Term Modeling (CVPR 2025 · similarity 0.87)
- UniMuMo: Unified Text, Music, and Motion Generation (AAAI 2025 · similarity 0.86)
- MuPT: A Generative Symbolic Music Pretrained Transformer (ICLR 2025 · similarity 0.86)
- Text2midi: Generating Symbolic Music from Captions (AAAI 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
