Research map / Detection and segmentation / Open-set and few-shot detection
Visually Indicated Sounds
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 spacetime images to RGB input (ablation) | Impact detection model · Impact detection | average precision: 21.6% -> 43.6% | Helped |
| Use the recurrent model instead of image matching | Sound prediction system · Psychophysical real-or-fake test | participant mistake rate | Helped |
| Add ImageNet pretraining (ablation) | Sound prediction model · Psychophysical real-or-fake test | participant mistake rate | No clear effect |
| Add spacetime images to RGB input (ablation) | Sound prediction model · Psychophysical real-or-fake test | participant mistake rate | No clear effect |
| Retain recurrent connections rather than resetting the hidden state between timesteps (ablation) | Sound prediction model · Sound prediction | automated sound metrics | Helped |
| Use example-based waveform generation instead of parametric inversion (ablation) | Sound prediction model · Psychophysical real-or-fake test | participant mistake rate | Mixed |
| Add ImageNet pretraining (ablation) | Sound prediction model · Material classification from predicted sounds | class-averaged accuracy: 22.7% -> 28.8% | Helped |
| Add ImageNet pretraining (ablation) | Sound prediction model · Reaction classification from predicted sounds | class-averaged accuracy: 53.5% -> 55.2% | Helped |
| Add ImageNet pretraining (ablation) | Sound prediction model · Action classification from predicted sounds | class-averaged accuracy: 68.6% -> 66.5% | Hurt |
| Use material labels to restrict example matches to the same material (ablation) | Example-based synthesis · Psychophysical real-or-fake test | participant mistake rate | Helped |
| Use predicted sounds instead of randomly sampling a sound from the same material class | Sound prediction system · Psychophysical real-or-fake test | participant mistake rate | Helped |
| Use predicted sounds instead of image-matched sounds | Sound prediction system · Loudness and spectral-centroid prediction | mean squared error | Helped |
Closest papers by meaning
- Shape and Material from Sound (NeurIPS 2017 · similarity 0.87)
- Hearing Hands: Generating Sounds from Physical Interactions in 3D Scenes (CVPR 2025 · similarity 0.86)
- Physics-Driven Diffusion Models for Impact Sound Synthesis From Videos (CVPR 2023 · similarity 0.85)
- Sounding that Object: Interactive Object-Aware Image to Audio Generation (ICML 2025 · similarity 0.82)
- Images that Sound: Composing Images and Sounds on a Single Canvas (NeurIPS 2024 · similarity 0.82)
- Generative Perception of Shape and Material from Differential Motion (NeurIPS 2025 · similarity 0.81)
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.