Research map / Video / Action recognition
Symbiotic Attention with Privileged Information for Egocentric Action Recognition
Xiaohan Wang, Yu Wu, Linchao Zhu, Yi Yang
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 SAP interactions among verb, noun, and privileged detection features | 3D ResNet-50 action recognition · EPIC-Kitchens validation | noun top-1 accuracy: 23.8 -> 35.0 | Improved |
| add SAP interactions among verb, noun, and privileged detection features | 3D ResNet-50 action recognition · EPIC-Kitchens validation | verb top-1 accuracy: 54.6 -> 55.9 | Improved |
| add SAP interactions among verb, noun, and privileged detection features | 3D ResNet-50 action recognition · EPIC-Kitchens validation | action top-1 accuracy: 25.0; 2.2% higher than LFB Max | Improved |
| use SAP rather than LFB Max | EPIC-Kitchens test seen (S1) · EPIC-Kitchens test seen (S1) | action top-1 accuracy: 34.8 vs 32.7; 2.1% | Improved |
| use SAP rather than LFB Max | EPIC-Kitchens test unseen (S2) · EPIC-Kitchens test unseen (S2) | action top-1 accuracy: 23.9 vs 21.2; 2.7% | Improved |
| use SAP rather than LSTA | EGTEA · EGTEA average | top-1 accuracy: 62.7 vs 61.9 | Improved |
| add verb features to noun-only CNN (ablation) | EPIC-Kitchens noun prediction · EPIC-Kitchens validation | noun top-1 accuracy: 23.8 -> 24.2 | Improved |
| use privileged detection features alone with average pooling (ablation) | EPIC-Kitchens noun prediction · EPIC-Kitchens validation | noun top-1 accuracy: 23.8 -> 24.5 | Improved |
| use privileged detection features alone with max pooling (ablation) | EPIC-Kitchens noun prediction · EPIC-Kitchens validation | noun top-1 accuracy: 23.8 -> 25.6 | Improved |
| add cross-stream gating (ablation) | SAP noun prediction · EPIC-Kitchens validation | noun top-1 accuracy: 30.4 -> 32.7 | Improved |
| add gating (ablation) | SAP noun prediction · EPIC-Kitchens validation | noun top-1 accuracy: 33.6 -> 35.0 | Improved |
| add cross-stream interaction instead of gating and attending with the same stream (ablation) | SAP noun prediction · EPIC-Kitchens validation | noun top-1 accuracy: 33.2 -> 35.0 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- LSTA — LSTA: Long Short-Term Attention for Egocentric Action Recognition (CVPR 2019) · better on 1 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
- Symbiotic Attention for Egocentric Action Recognition With Object-Centric Alignment (TPAMI 2020 · similarity 0.96)
- Opening the Vocabulary of Egocentric Actions (NeurIPS 2023 · similarity 0.90)
- Interactive Prototype Learning for Egocentric Action Recognition (ICCV 2021 · similarity 0.89)
- Action Scene Graphs for Long-Form Understanding of Egocentric Videos (CVPR 2024 · similarity 0.88)
- Do Egocentric Video-Language Models Truly Understand Hand-Object Interactions? (ICLR 2025 · similarity 0.87)
- EgoThinker: Unveiling Egocentric Reasoning with Spatio-Temporal CoT (NeurIPS 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.
