Research map / 3D vision / 3D generation
STCrowd: A Multimodal Dataset for Pedestrian Perception in Crowded Scenes
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 DHA to Voxel-Center | Voxel-Center LiDAR detection · STCrowd validation | mAP: 0.582 -> 0.617 | Helped |
| Add DHA to Voxel-Center | Voxel-Center 3D pedestrian detection · Waymo W30 level 1, W30 level 2, W50 level 1, W50 level 2, H3D, and nuScenes | mAP: 0.701 -> 0.726; 0.649 -> 0.673; 0.598 -> 0.638; 0.532 -> 0.569; 0.595 -> 0.609; 0.783 -> 0.795 | Helped |
| Add DHA to Voxel-Center tracking | Voxel-Center · STCrowd validation | MOTA: 0.342 -> 0.368 | Helped |
| Add DHA to Pillar-Center (ablation) | Pillar-Center LiDAR detection · STCrowd validation | mAP: 0.541 -> 0.591 | Helped |
| Add DHA to PointAugmenting (ablation) | PointAugmenting fusion detection · STCrowd validation | mAP: 0.587 -> 0.594 | Helped |
| Add the spatial attention module (ablation) | DHA on Voxel-Center · STCrowd validation | mAP: 0.601 -> 0.617 | Helped |
| Add hierarchical heatmaps (ablation) | DHA on Voxel-Center · STCrowd validation | mAP: 0.603 -> 0.617 | Helped |
| Use DHA rather than the anchor-based PointPillar baseline | STCrowd LiDAR detection · STCrowd validation | mAP: 0.245 -> 0.617 | Helped |
| Use image-LiDAR fusion rather than LiDAR-only Voxel-Center | STCrowd 3D detection · STCrowd validation | mAP: 0.582 -> 0.601 | Helped |
| Use DHA rather than Voxel-Center | STCrowd occlusion-level pedestrian recall · STCrowd validation | AR_0: 0.859 -> 0.902 | Helped |
Closest papers by meaning
- Tracking Pedestrian Heads in Dense Crowd (CVPR 2021 · similarity 0.86)
- Repulsion Loss: Detecting Pedestrians in a Crowd (CVPR 2018 · similarity 0.85)
- WILDTRACK: A Multi-Camera HD Dataset for Dense Unscripted Pedestrian Detection (CVPR 2018 · similarity 0.84)
- Multi-View Pedestrian Occupancy Prediction with a Novel Synthetic Dataset (AAAI 2025 · similarity 0.84)
- Point-to-Region Loss for Semi-Supervised Point-Based Crowd Counting (CVPR 2025 · similarity 0.83)
- GigaTraj: Predicting Long-term Trajectories of Hundreds of Pedestrians in Gigapixel Complex Scenes (CVPR 2024 · 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.