Research map / Detection and segmentation / CNN features
Occluded Pedestrian Detection Through Guided Attention in CNNs
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 |
|---|---|---|---|
| Add part-detection-guided channel attention | FasterRCNN · CityPersons val, HO | MR: 64.83 -> 56.66; +8.17 pp | Helped |
| Add visible-box-guided channel attention | FasterRCNN · CityPersons val, HO | MR: 64.83 -> 57.31; +7.52 pp | Helped |
| Use self-attention guidance (ablation) | FasterRCNN · CityPersons val, R | MR: 15.52 -> 20.93; -5.41 pp | Hurt |
| Use self-attention guidance (ablation) | FasterRCNN · CityPersons val, R+HO | MR: 41.45 -> 40.83; +0.62 pp | Helped |
| Use self-attention guidance (ablation) | FasterRCNN · CityPersons val across R and HO | MR: R: 15.52 -> 20.93; HO: 64.83 -> 58.33 | Mixed |
| Add part-detection-guided attention rather than using part heatmaps as direct classification features (ablation) | FasterRCNN · CityPersons val, R+HO | MR: 40.64 -> 38.23 | Helped |
| Add visible-box-guided attention | FasterRCNN · Caltech test, HO | MR: 57.58 -> 45.18 | Helped |
| Add part-detection-guided attention | FasterRCNN · ETH, R+HO | MR: 35.64 -> 33.84; +1.80 pp | Helped |
Closest papers by meaning
- Temporal-Context Enhanced Detection of Heavily Occluded Pedestrians (CVPR 2020 · similarity 0.88)
- Where, What, Whether: Multi-Modal Learning Meets Pedestrian Detection (CVPR 2020 · similarity 0.88)
- CityPersons: A Diverse Dataset for Pedestrian Detection (CVPR 2017 · similarity 0.87)
- What Can Help Pedestrian Detection? (CVPR 2017 · similarity 0.87)
- Attention Disturbance and Dual-Path Constraint Network for Occluded Person Re-identification (AAAI 2024 · similarity 0.85)
- Multi-View Pedestrian Occupancy Prediction with a Novel Synthetic Dataset (AAAI 2025 · similarity 0.83)
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.