Research map / Video / Video segmentation and tracking
Ego-Topo: Environment Affordances From Egocentric Video
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 |
|---|---|---|---|
| Functionally link zones across kitchens instead of using a single-video map | Ego-Topo affordance prediction · EPIC-Kitchens | mAP (All): 28.7 -> 29.4 | Helped |
| Functionally link zones across kitchens instead of using a single-video map | Ego-Topo affordance prediction · EPIC-Kitchens | mAP (Rare): 18.9 -> 19.7 | Helped |
| Link zones across videos of the same kitchen instead of using a single-video map | Ego-Topo affordance prediction · EGTEA+ | mAP (All): 48.9 -> 51.6 | Helped |
| Link zones across videos of the same kitchen instead of using a single-video map | Ego-Topo affordance prediction · EGTEA+ | mAP (Rare): 35.3 -> 37.8 | Helped |
| Add graph convolution to aggregate information from neighboring zones (ablation) | Ego-Topo long-term action anticipation · EPIC-Kitchens | mAP (All): 34.6 -> 38.0 | Helped |
| Add graph convolution to aggregate information from neighboring zones (ablation) | Ego-Topo long-term action anticipation · EGTEA+ | mAP (All): 72.5 -> 73.5 | Helped |
| Add graph convolution to aggregate information from neighboring zones (ablation) | Ego-Topo long-term action anticipation · EGTEA+ | mAP (Rare): 54.2 -> 54.7 | Helped |
| Use a visual-only clustering of action clips rather than activity-based topological zones | Affordance prediction · EPIC-Kitchens | mAP (All): 26.7 vs. 29.4 | Hurt |
| Use SLAM grid cells to link actions by location rather than discovered zones | Affordance prediction · EPIC-Kitchens | mAP (All): 26.6 vs. 29.4 | Hurt |
| Use ActionMaps location and feature similarity linking rather than Ego-Topo zones | Affordance prediction · EPIC-Kitchens | mAP (All): 21.0 vs. 29.4 | Hurt |
| Use Ego-Topo rather than Timeception for long-term action anticipation | Long-term action anticipation · EPIC-Kitchens | mAP (All): 35.6 -> 38.0 | Helped |
| Use Ego-Topo rather than Timeception for long-term action anticipation | Long-term action anticipation · EGTEA+ | mAP (All): 74.1 vs. 73.5 | Hurt |
Closest papers by meaning
- Weakly Supervised Multimodal Affordance Grounding for Egocentric Images (AAAI 2024 · similarity 0.87)
- EgoChoir: Capturing 3D Human-Object Interaction Regions from Egocentric Views (NeurIPS 2024 · similarity 0.87)
- Learning to Evolve Scenes: Reasoning about Human Activities with Scene Graphs (NeurIPS 2026 · similarity 0.87)
- DIV-FF: Dynamic Image-Video Feature Fields For Environment Understanding in Egocentric Videos (CVPR 2025 · similarity 0.86)
- SAW-Bench: Learning Situated Awareness in the Real World (ICML 2026 · similarity 0.86)
- EgoGen: An Egocentric Synthetic Data Generator (CVPR 2024 · 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.