Research map / Video / Video generation
Sound2Sight: Generating Visual Dynamics from Sound and Context
Moitreya Chatterjee, Anoop Cherian
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 |
|---|---|---|---|
| add AA loss term to the discriminator (ablation) | Sound2Sight · M3SO-NB frame 25 | SSIM: 0.8296 -> 0.8697 | Improved |
| add MD loss term to the discriminator (ablation) | Sound2Sight · M3SO-NB frame 25 | SSIM: 0.8251 -> 0.8296 | Improved |
| add AA loss term to the discriminator (ablation) | Sound2Sight · YouTube Painting frame 45 | SSIM: 0.9076 -> 0.9110 | Improved |
| add MD loss term to the discriminator (ablation) | Sound2Sight · YouTube Painting frame 45 | SSIM: 0.9022 -> 0.9076 | Improved |
| add audio input (ablation) | Sound2Sight · M3SO-NB frame 25 | SSIM: 0.6920 -> 0.8697 | Improved |
| add audio input (ablation) | Sound2Sight · YouTube Painting frame 45 | SSIM: 0.8986 -> 0.9110 | Improved |
| add posterior network and multimodal discriminator over L2-only training (ablation) | Sound2Sight · M3SO-NB frame 25 | SSIM: 0.5610 -> 0.8697 | Improved |
| add teacher forcing to training (ablation) | Sound2Sight · M3SO-NB frame 25 | SSIM: 0.8697 -> 0.8519 | Hurt |
| add teacher forcing to training (ablation) | Sound2Sight · YouTube Painting frame 45 | SSIM: 0.9110 -> 0.9109 | No clear effect |
| use mismatched audio and past frames at inference (ablation) | Sound2Sight · M3SO-NB frame 25 | SSIM: 0.8697 -> 0.8234 | Hurt |
| use transformer encoder instead of LSTM (ablation) | Sound2Sight · M3SO-NB | SSIM | Improved |
| sample more candidate futures per time step (1 to 100) | Sound2Sight · M3SO-NB | SSIM of best-matching sample | Improved |
Later papers that compared with this method
4 later accepted papers report Sound2Sight as a baseline in their results tables; 3 of them report a better number on most of the results they share with it.
- The Power of Sound (TPoS): Audio Reactive Video Generation with Stable Diffusion (ICCV 2023 · better on 2 of 2 shared results)
- Sound-Guided Semantic Video Generation (ECCV 2022 · better on 6 of 6 shared results)
- Long Video Generation with Time-Agnostic VQGAN and Time-Sensitive Transformer (ECCV 2022 · better on 2 of 2 shared results)
- CCVS: Context-aware Controllable Video Synthesis (NeurIPS 2021 · better on 0 of 1 shared result)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- From Vision to Audio and Beyond: A Unified Model for Audio-Visual Representation and Generation (ICML 2024 · similarity 0.89)
- PrismAudio: Decomposed Chain-of-Thought and Multi-dimensional Rewards for Video-to-Audio Generation (ICLR 2026 · similarity 0.88)
- V2A-Mapper: A Lightweight Solution for Vision-to-Audio Generation by Connecting Foundation Models (AAAI 2024 · similarity 0.88)
- Seeing and Hearing: Open-domain Visual-Audio Generation with Diffusion Latent Aligners (CVPR 2024 · similarity 0.88)
- CASP: Consistency-aware Audio-induced Saliency Prediction Model for Omnidirectional Video (CVPR 2025 · similarity 0.88)
- A Versatile Diffusion Transformer with Mixture of Noise Levels for Audiovisual Generation (NeurIPS 2024 · similarity 0.87)
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.
