Research map / Video / Video segmentation and tracking
PANDA: A Gigapixel-Level Human-Centric Video Dataset
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 local interaction clips selected randomly to global trajectory features (ablation) | PANDA group detection · PANDA group detection | F1: 0.160 -> 0.172 | Helped |
| Select local interaction clips using uncertainty instead of random selection (ablation) | PANDA group detection with global trajectories and local clips · PANDA group detection | F1: 0.172 -> 0.207 | Helped |
| Increase the number of local clips selected for zoom-in (ablation) | PANDA group detection · PANDA group detection | performance | Helped |
| Increase dropout sampling count from 10 to 510 (ablation) | PANDA group detection uncertainty estimation · PANDA group detection | performance | No clear effect |
| Use RetinaNet rather than Faster R-CNN as the detector | PANDA tracking · PANDA tracking | MOTA: 25.53 -> 16.36 | Hurt |
| Use Faster R-CNN rather than RetinaNet as the detector | PANDA detection · PANDA visible body, small | AP_.50: 0.171 -> 0.201 | Helped |
Closest papers by meaning
- GigaTraj: Predicting Long-term Trajectories of Hundreds of Pedestrians in Gigapixel Complex Scenes (CVPR 2024 · similarity 0.86)
- MVHumanNet: A Large-scale Dataset of Multi-view Daily Dressing Human Captures (CVPR 2024 · similarity 0.86)
- HOI-M^3: Capture Multiple Humans and Objects Interaction within Contextual Environment (CVPR 2024 · similarity 0.85)
- GigaHumanDet: Exploring Full-Body Detection on Gigapixel-Level Images (AAAI 2024 · similarity 0.85)
- Leader360V: A Large-scale, Real-world 360 Video Dataset for Multi-task Learning in Diverse Environment (NeurIPS 2025 · similarity 0.85)
- HARDVS: Revisiting Human Activity Recognition with Dynamic Vision Sensors (AAAI 2024 · similarity 0.84)
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.