Pose Estimation On Coco Val2017
Metrics
AP
AR
Results
Performance results of various models on this benchmark
Model Name | AP | AR | Paper Title | Repository |
|---|---|---|---|---|
| MogaNet-S (256x192) | 74.9 | 80.1 | MogaNet: Multi-order Gated Aggregation Network | |
| MogaNet-T (256x192) | 73.2 | 78.8 | MogaNet: Multi-order Gated Aggregation Network | |
| ViTPose-B (Single-task_GT-bbox_256x192) | 77.3 | 80.4 | ViTPose: Simple Vision Transformer Baselines for Human Pose Estimation | |
| Bias (HRNet_256x192) | 75.8 | - | Removing the Bias of Integral Pose Regression | - |
| HRNet (256x192) | 75.3 | - | Deep High-Resolution Representation Learning for Human Pose Estimation | |
| MogaNet-B (384x288) | 77.3 | 82.2 | MogaNet: Multi-order Gated Aggregation Network | |
| SimpleBaseLine (256x192) | 70.4 | - | Simple Baselines for Human Pose Estimation and Tracking | |
| MogaNet-S (384x288) | 76.4 | 81.4 | MogaNet: Multi-order Gated Aggregation Network | |
| CCNet (ViTPose-B_GT-bbox_256x192) | 78.1 | 80.4 | On the Calibration of Human Pose Estimation | - |
| ViTPose-B (Single-task_Det-bbox_256x192) | 75.8 | 81.1 | ViTPose: Simple Vision Transformer Baselines for Human Pose Estimation | |
| RLE (256x192) | 71.3 | - | Human Pose Regression with Residual Log-likelihood Estimation |
0 of 11 row(s) selected.