Research map / Detection and segmentation / Tracking
Multi-Person Articulated Tracking With Spatial and Temporal Embeddings
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 |
|---|---|---|---|
| Combine KE and SIE rather than use either embedding alone (ablation) | SpatialNet · ICCV'17 PoseTrack validation | AP (Total): 70.9%/71.3% -> 74.0% | Helped |
| Add geometric ordinal auxiliary training to KE+SIE (ablation) | SpatialNet · ICCV'17 PoseTrack validation | AP (Total): 74.0 -> 74.3 | Helped |
| Add PGG to KE (ablation) | SpatialNet · ICCV'17 PoseTrack validation | AP (Total): 70.9 -> 73.2 | Helped |
| Add PGG to SIE (ablation) | SpatialNet · ICCV'17 PoseTrack validation | AP (Total): 71.3 -> 75.1 | Helped |
| Add PGG to combined KE+SIE+aux (ablation) | SpatialNet · ICCV'17 PoseTrack validation | AP (Total): 74.3 -> 77.0 | Helped |
| Add HE and TIE together instead of using KE+SIE baseline tracking (ablation) | Pose tracker · ICCV'17 PoseTrack validation | MOTA (Total): 67.7 -> 69.2 | Helped |
| Add spatial-temporal PGG to HE+TIE (ablation) | TemporalNet · ICCV'17 PoseTrack validation | MOTA (Total): 69.2% vs. 71.8% | Helped |
| Use multi-scale rather than single-scale inference (ablation) | SpatialNet · MS-COCO subset | AP: 0.650 -> 0.680 | Helped |
| Use the proposed HE+TIE+PGG tracker rather than FlowTrack | Pose tracking · ICCV'17 PoseTrack validation | MOTA (Total): 65.4% to 71.8% | Helped |
Closest papers by meaning
- ArtTrack: Articulated Multi-Person Tracking in the Wild (CVPR 2017 · similarity 0.92)
- PoseTrack: Joint Multi-Person Pose Estimation and Tracking (CVPR 2017 · similarity 0.90)
- Combining Detection and Tracking for Human Pose Estimation in Videos (CVPR 2020 · similarity 0.90)
- PoseTrack: A Benchmark for Human Pose Estimation and Tracking (CVPR 2018 · similarity 0.89)
- CoMotion: Concurrent Multi-person 3D Motion (ICLR 2025 · similarity 0.88)
- Motions as Queries: One-Stage Multi-Person Holistic Human Motion Capture (CVPR 2025 · similarity 0.88)
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.