Pose Tracking On Posetrack2017
Metrics
MOTA
mAP
Results
Performance results of various models on this benchmark
| Paper Title | |||
|---|---|---|---|
| DetTrack | 64.09 | 74.14 | Combining detection and tracking for human pose estimation in videos |
| KeyTrack | 61.15 | 74.04 | 15 Keypoints Is All You Need |
| LightTrack | 58.01 | 66.65 | LightTrack: A Generic Framework for Online Top-Down Human Pose Tracking |
| HRNet-W48 COCO | 57.93 | 74.95 | Deep High-Resolution Representation Learning for Human Pose Estimation |
| MSRA (FlowTrack) | 57.81 | 74.57 | Simple Baselines for Human Pose Estimation and Tracking |
| TML++ (MIPAL) | 54.46 | 68.78 | Pose estimator and tracker using temporal flow maps for limbs |
| STAF | 53.81 | 70.28 | Efficient Online Multi-Person 2D Pose Tracking with Recurrent Spatio-Temporal Affinity Fields |
| ProTracker | 51.82 | 59.56 | Detect-and-Track: Efficient Pose Estimation in Videos |
| PoseFlow | 50.98 | 62.95 | Pose Flow: Efficient Online Pose Tracking |
| PoseTrack | 48.37 | 59.22 | PoseTrack: A Benchmark for Human Pose Estimation and Tracking |
0 of 10 row(s) selected.