Research map / Video / Video segmentation and tracking
Integrating Egocentric Videos in Top-view Surveillance Videos: Joint Identification and Temporal Alignment
Shervin Ardeshir, Ali Borji
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 geometric and spatiotemporal reasoning (fusion) on top of the two-stream supervised visual re-id (ablation) | Ours (supervised) · Cross-view re-identification | AUC of CMC: 0.668 -> 0.716 | Improved |
| add geometric and spatiotemporal reasoning (fusion) on top of unsupervised VGG-19 visual re-id (ablation) | Ours (unsupervised) · Cross-view re-identification | AUC of CMC: 0.537 -> 0.612 | Improved |
| add fusion with geometric and spatiotemporal reasoning to Ahmed's re-id network (ablation) | Ahmed [21] · Cross-view re-identification | AUC of CMC: 0.563 -> 0.621 | Improved |
| add fusion with geometric and spatiotemporal reasoning to Cheng's re-id network (ablation) | Cheng [22] · Cross-view re-identification | AUC of CMC: 0.581 -> 0.634 | Improved |
| replace the unsupervised visual module with the supervised two-stream network (before fusion) (ablation) | visual reasoning · Cross-view re-identification | AUC of CMC: 0.537 -> 0.668 | Improved |
| estimate time offset with unsupervised joint method instead of assuming tau=0 | temporal alignment · Time-delay estimation | mean error (frames): 186.5 -> 138.9 frames | Improved |
| add visual supervision (two-stream network) to time-delay estimation (ablation) | temporal alignment · Time-delay estimation | mean error (frames): 138.9 -> 120.6 frames | Improved |
| provide ground-truth self-identity to time-delay estimation (ablation) | temporal alignment · Time-delay estimation | mean error (frames): 120.6 -> 97.39 frames | Improved |
| additionally provide ground-truth re-identification labels once self-identity is known (ablation) | temporal alignment · Time-delay estimation | mean error (frames): 97.39 -> 90.32 frames | Improved |
Closest papers by meaning
- Egocentric Meets Top-View (TPAMI 2018 · similarity 0.93)
- Ego2Top: Matching Viewers in Egocentric and Top-View Videos (ECCV 2016 · similarity 0.93)
- EgoSim: An Egocentric Multi-view Simulator and Real Dataset for Body-worn Cameras during Motion and Activity (NeurIPS 2024 · similarity 0.86)
- Fusing Personal and Environmental Cues for Identification and Segmentation of First-Person Camera Wearers in Third-Person Views (CVPR 2024 · similarity 0.86)
- Instance Tracking in 3D Scenes from Egocentric Videos (CVPR 2024 · similarity 0.85)
- Test-Time Adaptation for Combating Missing Modalities in Egocentric Videos (ICLR 2025 · similarity 0.85)
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.
