Multi-Person Pose Estimation
Multi-person pose estimation is the task of estimating the pose of multiple people in one frame.
( Image credit: Human Pose Estimation with TensorFlow )
Papers
Showing 1–10 of 151 papers
All datasetsCrowdPoseCOCO (Common Objects in Context)COCO test-devMPII Multi-PersonOCHumanCOCO minivalPoseTrack2018PoseTrack2017WAFCOCOMulti-Person PoseTrackPoseTrack21
Benchmark Results
| # | Model | Metric | Claimed | Verified | Status |
|---|---|---|---|---|---|
| 1 | RTMO-l | mAP @0.5:0.95 | 83.8 | — | Unverified |
| 2 | BUCTD-W48 (w/cond. input from PETR, and generative sampling) | mAP @0.5:0.95 | 78.5 | — | Unverified |
| 3 | I²R-Net (1st stage: HRFormer-B) | mAP @0.5:0.95 | 77.4 | — | Unverified |
| 4 | ED-Pose (Swin-L) | mAP @0.5:0.95 | 76.6 | — | Unverified |
| 5 | DETRPose-X | mAP @0.5:0.95 | 75.1 | — | Unverified |
| 6 | DETRPose-L | mAP @0.5:0.95 | 73.3 | — | Unverified |
| 7 | HRFormer-B | mAP @0.5:0.95 | 72.4 | — | Unverified |
| 8 | BAPose (W32) | mAP @0.5:0.95 | 72.2 | — | Unverified |
| 9 | DETRPose-M | mAP @0.5:0.95 | 72 | — | Unverified |
| 10 | TransPose-H | mAP @0.5:0.95 | 71.8 | — | Unverified |
| # | Model | Metric | Claimed | Verified | Status |
|---|---|---|---|---|---|
| 1 | EvoPose2D-L | Test AP | 76.8 | — | Unverified |
| 2 | PoseFix | Test AP | 76.7 | — | Unverified |
| 3 | LitePose-S | Test AP | 56.7 | — | Unverified |
| 4 | RSN | AP | 0.79 | — | Unverified |
| 5 | DarkPose | AP | 0.77 | — | Unverified |
| 6 | UniPose | AP | 0.77 | — | Unverified |
| 7 | CPN+ | AP | 0.73 | — | Unverified |
| 8 | BAPose | AP | 0.73 | — | Unverified |
| 9 | CenterGroup | AP | 0.71 | — | Unverified |
| 10 | OpenPifPaf | AP | 0.71 | — | Unverified |
| # | Model | Metric | Claimed | Verified | Status |
|---|---|---|---|---|---|
| 1 | SCIO (HRNet-48) | AP | 79.2 | — | Unverified |
| 2 | HRNet-W48plus | AP | 78.7 | — | Unverified |
| 3 | HRNet-W32 | AP | 76.2 | — | Unverified |
| 4 | ResNet50 | AP | 73.7 | — | Unverified |
| 5 | HigherHRNet (ScaleNet_P4) | AP | 71.6 | — | Unverified |
| 6 | HigherHRNet (HR-Net-48) | AP | 70.5 | — | Unverified |
| 7 | SMPR (HR-Net-32) | AP | 70.2 | — | Unverified |
| 8 | PersonLab | AP | 68.7 | — | Unverified |
| 9 | Identity Mapping Hourglass | AP | 68.1 | — | Unverified |
| 10 | SPM | AP | 66.9 | — | Unverified |
| # | Model | Metric | Claimed | Verified | Status |
|---|---|---|---|---|---|
| 1 | AlphaPose | AP | 82.1 | — | Unverified |
| 2 | Generative Partition Networks | AP | 80.4 | — | Unverified |
| 3 | SPM | AP | 78.5 | — | Unverified |
| 4 | Refine | AP | 78 | — | Unverified |
| 5 | Associative Embedding | AP | 77.5 | — | Unverified |
| 6 | Part Affinity Fields | AP | 75.6 | — | Unverified |
| 7 | Articulated Tracking | AP | 74.3 | — | Unverified |
| 8 | Local Joint-to-Person Association | AP | 62.2 | — | Unverified |
| 9 | DeeperCut | AP | 59.4 | — | Unverified |
| # | Model | Metric | Claimed | Verified | Status |
|---|---|---|---|---|---|
| 1 | MIPNet (gt-bb) | AP50 | 89.7 | — | Unverified |
| 2 | I²R-Net (1st stage:TransPose-H) | AP50 | 85 | — | Unverified |
| 3 | TransPose-H | AP50 | 82.7 | — | Unverified |
| 4 | HRFormer-B | AP50 | 81.4 | — | Unverified |
| 5 | SPM | AP50 | 67.5 | — | Unverified |
| 6 | CrowdPose | AP50 | 40.8 | — | Unverified |
| 7 | SimplePose | AP50 | 37.4 | — | Unverified |
| 8 | Mask R-CNN | AP50 | 33.2 | — | Unverified |
| # | Model | Metric | Claimed | Verified | Status |
|---|---|---|---|---|---|
| 1 | HRNet-W48plus | AP | 79.1 | — | Unverified |
| 2 | HRNet-W32 | AP | 77.8 | — | Unverified |
| 3 | ResNet50 | AP | 75.3 | — | Unverified |
| 4 | InsPose | AP | 63.1 | — | Unverified |
| # | Model | Metric | Claimed | Verified | Status |
|---|---|---|---|---|---|
| 1 | Poseidon | Mean mAP | 87.8 | — | Unverified |
| 2 | DCPose | Mean mAP | 79 | — | Unverified |
| 3 | PoseWarper | Mean mAP | 78 | — | Unverified |
| 4 | Refine | Mean mAP | 73.8 | — | Unverified |
| # | Model | Metric | Claimed | Verified | Status |
|---|---|---|---|---|---|
| 1 | DCPose | Mean mAP | 79.2 | — | Unverified |
| 2 | PoseWarper | Mean mAP | 77.94 | — | Unverified |
| 3 | PoseTrack | Mean mAP | 59.4 | — | Unverified |
| # | Model | Metric | Claimed | Verified | Status |
|---|---|---|---|---|---|
| 1 | DeeperCut | AOP | 88.1 | — | Unverified |
| 2 | DeepCut | AOP | 86.5 | — | Unverified |
| 3 | Generative Partition Networks | AP | 84.8 | — | Unverified |
| # | Model | Metric | Claimed | Verified | Status |
|---|---|---|---|---|---|
| 1 | CMU-Pose | AP | 0.62 | — | Unverified |
| # | Model | Metric | Claimed | Verified | Status |
|---|---|---|---|---|---|
| 1 | PoseTrack | Mean mAP | 38.2 | — | Unverified |
| # | Model | Metric | Claimed | Verified | Status |
|---|---|---|---|---|---|
| 1 | Poseidon | Mean mAP | 88.3 | — | Unverified |