Research map / Video / Video-language understanding
Egocentric Video Task Translation
Read the paperCheck 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 |
|---|---|---|---|
| translate auxiliary-task features with EgoT2-s | task-specific models · OSCC | accuracy (%): 68.22 -> 72.69 | Helped |
| translate auxiliary-task features with EgoT2-s | task-specific models · TTM | mAP (%): 58.91 -> 66.54 | Helped |
| translate auxiliary-task features with EgoT2-s | task-specific models · ASD | mAP (%): 79.05 -> 79.38 | Helped |
| use one task-general translator for all tasks | EgoT2-g · TTM | mAP (%): 58.91 -> 64.49 | Helped |
| replace the TTM auxiliary-task models with TTM-only models (ablation) | EgoT2-s · TTM | mAP (%): 63.40 -> 66.54 | Helped |
| preserve temporal information instead of pooling task features (ablation) | EgoT2-s · TTM | mAP (%): 65.47 -> 66.54 | Helped |
| freeze task-specific models during second-stage training (ablation) | EgoT2-s · TTM | mAP (%): 66.00 -> 66.54 | No clear effect |
| add auxiliary-task models rather than use OSCC-only models (ablation) | EgoT2-s · OSCC | mAP (%): 69.68 -> 72.69 | Helped |
| preserve temporal information instead of pooling task features (ablation) | EgoT2-s · OSCC | mAP (%): 71.65 -> 72.69 | Helped |
| freeze task-specific models during second-stage training (ablation) | EgoT2-s · OSCC | mAP (%): 72.22 -> 72.69 | No clear effect |
| replace the SlowFast backbone with EgoVLP while retaining task translation | EgoT2-s · OSCC | accuracy (%): 73.00 -> 75.77 | Helped |
| include LAM from the other task cluster | EgoT2-g · PNR | temporal localization error (s): 0.612 -> 0.611 | No clear effect |
Closest papers by meaning
- A Backpack Full of Skills: Egocentric Video Understanding with Diverse Task Perspectives (CVPR 2024 · similarity 0.88)
- Ego-Exo4D: Understanding Skilled Human Activity from First- and Third-Person Perspectives (CVPR 2024 · similarity 0.88)
- Retrieval-Augmented Egocentric Video Captioning (CVPR 2024 · similarity 0.87)
- Action Scene Graphs for Long-Form Understanding of Egocentric Videos (CVPR 2024 · similarity 0.87)
- EgoThinker: Unveiling Egocentric Reasoning with Spatio-Temporal CoT (NeurIPS 2025 · similarity 0.87)
- Do Egocentric Video-Language Models Truly Understand Hand-Object Interactions? (ICLR 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.