Research map / Detection and segmentation / CNN features
Temporal-Context Enhanced Detection of Heavily Occluded Pedestrians
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 |
|---|---|---|---|
| use tube feature aggregation with local temporal context | Faster R-CNN pedestrian detector · Caltech HO | MR−2: 43.1 -> 30.9 | Helped |
| use tube feature aggregation with local temporal context | Faster R-CNN pedestrian detector · NightOwls Occluded | MR−2: 46.5 -> 42.1 | Helped |
| use tube feature aggregation with local temporal context | Faster R-CNN pedestrian detector · KAIST Heavy Occlusion | MR−2: 76.6 -> 71.3 | Helped |
| add the TDEM discriminative loss (ablation) | TFAN preliminary · Caltech HO | MR−2: 37.6 -> 32.7 | Helped |
| add spatial similarity to tube linking (ablation) | TFAN+TDEM · Caltech HO | MR−2: 33.5 -> 32.7 | Helped |
| add PRM to focus aggregation on visible parts (ablation) | TFAN+TDEM · Caltech HO | MR−2: 32.7 -> 30.9 | Helped |
| use adaptive weights rather than average weights (ablation) | TFAN · Caltech HO | MR−2: 35.0 -> 30.9 | Helped |
| vary tube length from τ = 3 to τ = 10 (ablation) | TFAN+TDEM+PRM · Caltech HO | MR−2: 32.9; 32.2; 31.5; 30.9; 31.1; 31.2; 31.5; 31.6 | Mixed |
Closest papers by meaning
- Occluded Pedestrian Detection Through Guided Attention in CNNs (CVPR 2018 · similarity 0.88)
- Improving Occlusion and Hard Negative Handling for Single-Stage Pedestrian Detectors (CVPR 2018 · similarity 0.86)
- PedHunter: Occlusion Robust Pedestrian Detector in Crowded Scenes (AAAI 2020 · similarity 0.86)
- Localized Semantic Feature Mixers for Efficient Pedestrian Detection in Autonomous Driving (CVPR 2023 · similarity 0.85)
- Where, What, Whether: Multi-Modal Learning Meets Pedestrian Detection (CVPR 2020 · similarity 0.85)
- Multi-View Pedestrian Occupancy Prediction with a Novel Synthetic Dataset (AAAI 2025 · 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.