Human Pose Estimation on LSP (test)
98.6Head AccuracyASR+AHO
Evaluation results
| Method | Links | |||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ASR+AHObackbone=Stacked HGs(8), augmentation=Adversarial Scaling and Rotating + Adversarial Hierarchical Occluding2018.05 | 98.6 | 95.3 | 92.8 | 90 | 94.8 | 95.3 | 94.5 | 94.5 | — | — | — | — | — | — | — | — | — | |
| Peng et al., CVPR'18# Param=26M, Deployment Cost=55G2018.11 | 98.6 | 95.3 | 92.8 | 90 | 94.8 | 95.3 | 94.5 | — | — | — | — | — | — | — | 94.5 | — | — | |
| Adversarial PoseNet2017.04 | 98.5 | 94 | 89.8 | 87.5 | 93.9 | 94.1 | 93 | 93.1 | — | — | — | — | — | — | — | — | — | |
| Stacked HGs(8)stacks=82018.05 | 98.2 | 94 | 91.2 | 87.2 | 93.5 | 94.5 | 92.6 | 93 | — | — | — | — | — | — | — | — | — | |
| Knowledge-Guided Deep Fractal Neural NetworksMetric=PCK@0.22017.05 | 98.2 | 94.4 | 91.8 | 89.3 | 94.7 | 95 | 93.5 | 93.9 | — | — | — | — | — | — | — | — | — | |
| Self Adversarial Training2017.07 | 98.2 | 94.9 | 92.2 | 89.5 | 94.2 | 95 | 94.1 | 94 | — | — | — | — | — | — | — | — | — | |
| Newell et al.publication=ECCV'162018.08 | 98.2 | 94 | 91.2 | 87.2 | 93.5 | 94.5 | 92.6 | 93 | — | — | — | — | — | — | — | — | — | |
| Chu et al.2018.05 | 98.1 | 93.7 | 89.3 | 86.9 | 93.4 | 94 | 92.5 | 92.6 | — | — | — | — | — | — | — | — | — | |
| Chu et al.2017.07 | 98.1 | 93.7 | 89.3 | 86.9 | 93.4 | 94 | 92.5 | 92.6 | — | — | — | — | — | — | — | — | — | |
| Chu et al. [13]Method Category=Detection based, Evaluation Protocol=Person-Centric (PC), PCK Threshold=0.22017.10 | 98.1 | 93.7 | 89.3 | 86.9 | 93.4 | 94 | 92.5 | — | — | — | — | — | — | 92.6 | — | — | — | |
| Chu et al.publication=CVPR'172018.08 | 98.1 | 93.7 | 89.3 | 86.9 | 93.4 | 94 | 92.5 | 92.6 | — | — | — | — | — | — | — | — | — | |
| [22]2017.04 | 97.8 | 92.5 | 87 | 83.9 | 91.5 | 89.9 | 87.2 | 90.1 | — | — | — | — | — | — | — | — | — | |
| [33]2017.04 | 97.8 | 92.5 | 87 | 83.9 | 91.5 | 90.8 | 89.9 | 90.5 | — | — | — | — | — | — | — | — | — | |
| Wei et al.2018.05 | 97.8 | 92.5 | 87 | 83.9 | 91.5 | 90.8 | 89.9 | 90.5 | — | — | — | — | — | — | — | — | — | |
| Wei et al.Metric=PCK@0.22017.05 | 97.8 | 92.5 | 87 | 83.9 | 91.5 | 90.8 | 89.9 | 90.5 | — | — | — | — | — | — | — | — | — | |
| Lifshitz et al.2017.07 | 97.8 | 93.3 | 85.7 | 80.4 | 85.3 | 76.6 | 70.2 | 85 | — | — | — | — | — | — | — | — | — | |
| Wei et al.2017.07 | 97.8 | 95 | 88.7 | 84 | 88.4 | 82.8 | 79.4 | 88.5 | — | — | — | — | — | — | — | — | — | |
| Wei et al. [43]Method Category=Detection based, Evaluation Protocol=Person-Centric (PC), PCK Threshold=0.22017.10 | 97.8 | 92.5 | 87 | 83.9 | 91.5 | 90.8 | 89.9 | — | — | — | — | — | — | 90.5 | — | — | — | |
| Wei et al.publication=CVPR'162018.08 | 97.8 | 92.5 | 87 | 83.9 | 91.5 | 90.8 | 89.9 | 90.5 | — | — | — | — | — | — | — | — | — | |
| Wei et al., CVPR'162016.09 | 97.8 | 92.5 | 87 | 83.9 | 91.5 | 90.8 | 89.9 | 90.5 | — | — | — | — | — | — | — | — | — | |
| Soft-argmaxMethod Category=Regression based, Evaluation Protocol=Person-Centric (PC), PCK Threshold=0.22017.10 | 97.5 | 93.3 | 87.6 | 84.6 | 92.8 | 92 | 90 | — | — | — | — | — | — | 91.1 | — | — | — | |
| [14]2017.04 | 97.4 | 92.7 | 87.5 | 84.4 | 91.5 | 89.9 | 87.2 | 90.1 | — | — | — | — | — | — | — | — | — | |
| Insafutdinov et al.2018.05 | 97.4 | 92.7 | 87.5 | 84.4 | 91.5 | 89.9 | 87.2 | 90.1 | — | — | — | — | — | — | — | — | — | |
| Insafutdinov et al.Metric=PCK@0.22017.05 | 97.4 | 92.7 | 87.5 | 84.4 | 91.5 | 89.9 | 87.2 | 90.1 | — | — | — | — | — | — | — | — | — | |
| Insafutdinov et al. [21]Method Category=Detection based, Evaluation Protocol=Person-Centric (PC), PCK Threshold=0.22017.10 | 97.4 | 92.7 | 87.5 | 84.4 | 91.5 | 89.9 | 87.2 | — | — | — | — | — | — | 90.1 | — | — | — | |
| Insafutdinov et al.publication=ECCV’162018.08 | 97.4 | 92.7 | 87.5 | 84.4 | 91.5 | 89.9 | 87.2 | 90.1 | — | — | — | — | — | — | — | — | — | |
| Insafutdinov et al., arXiv'162016.09 | 97.4 | 92.7 | 87.5 | 84.4 | 91.5 | 89.9 | 87.2 | 90.1 | — | — | — | — | — | — | — | — | — | |
| FPD# Param=3M, Deployment Cost=9G2018.11 | 97.3 | 92.3 | 86.8 | 84.2 | 91.9 | 92.2 | 90.9 | — | — | — | — | — | — | — | 90.8 | 64.3 | — | |
| [4]2017.04 | 97.2 | 92.1 | 88.1 | 85.2 | 92.2 | 91.4 | 88.7 | 90.7 | — | — | — | — | — | — | — | — | — | |
| Bulat et al.2018.05 | 97.2 | 92.1 | 88.1 | 85.2 | 92.2 | 91.4 | 88.7 | 90.7 | — | — | — | — | — | — | — | — | — | |
| Bulat & TzimiropoulosMetric=PCK@0.22017.05 | 97.2 | 92.1 | 88.1 | 85.2 | 92.2 | 91.4 | 88.7 | 90.7 | — | — | — | — | — | — | — | — | — | |
| Bulat et al.2017.07 | 97.2 | 92.1 | 88.1 | 85.2 | 92.2 | 91.4 | 88.7 | 90.7 | — | — | — | — | — | — | — | — | — | |
| Bulat and Tzimi. [5]Method Category=Detection based, Evaluation Protocol=Person-Centric (PC), PCK Threshold=0.22017.10 | 97.2 | 92.1 | 88.1 | 85.2 | 92.2 | 91.4 | 88.7 | — | — | — | — | — | — | 90.7 | — | — | — | |
| Bulat et al.publication=ECCV'162018.08 | 97.2 | 92.1 | 88.1 | 85.2 | 92.2 | 91.4 | 88.7 | 90.7 | — | — | — | — | — | — | — | — | — | |
| CU-Net-8Stacked U-Nets=82018.08 | 97.1 | 94.7 | 91.6 | 89 | 93.7 | 94.2 | 93.7 | 93.4 | — | — | — | — | — | — | — | — | — | |
| [21]2017.04 | 97 | 91 | 83.8 | 78.1 | 91 | 86.7 | 82 | 87.1 | — | — | — | — | — | — | — | — | — | |
| Pishchulin et al.2018.05 | 97 | 91 | 83.8 | 78.1 | 91 | 86.7 | 82 | 87.1 | — | — | — | — | — | — | — | — | — | |
| Pishchulin et al. CVPR'16Metric=PCK@0.22017.05 | 97 | 91 | 83.8 | 78.1 | 91 | 86.7 | 82 | 87.1 | — | — | — | — | — | — | — | — | — | |
| Pishchulin et al.2017.07 | 97 | 91 | 83.8 | 78.1 | 91 | 86.7 | 82 | 87.1 | — | — | — | — | — | — | — | — | — | |
| Pishchulin et al. [33]Method Category=Detection based, Evaluation Protocol=Person-Centric (PC), PCK Threshold=0.22017.10 | 97 | 91 | 83.8 | 78.1 | 91 | 86.7 | 82 | — | — | — | — | — | — | 87.1 | — | — | — | |
| Pishchulin et al.publication=CVPR'162018.08 | 97 | 91 | 83.8 | 78.1 | 91 | 86.7 | 82 | 87.1 | — | — | — | — | — | — | — | — | — | |
| Pishchulin et al., CVPR'162016.09 | 97 | 91 | 83.8 | 78.1 | 91 | 86.7 | 82 | 87.1 | — | — | — | — | — | — | — | — | — | |
| [16]2017.04 | 96.8 | 89 | 82.7 | 79.1 | 90.9 | 86 | 82.5 | 86.7 | — | — | — | — | — | — | — | — | — | |
| Lifshitz et al.2018.05 | 96.8 | 89 | 82.7 | 79.1 | 90.9 | 86 | 82.5 | 86.7 | — | — | — | — | — | — | — | — | — | |
| Lifshitz et al.Metric=PCK@0.22017.05 | 96.8 | 89 | 82.7 | 79.1 | 90.9 | 86 | 82.5 | 86.7 | — | — | — | — | — | — | — | — | — | |
| Insafutdinov et al.2017.07 | 96.8 | 95.2 | 89.3 | 84.4 | 88.4 | 83.4 | 78 | 88.5 | — | — | — | — | — | — | — | — | — | |
| Lifshitz et al. [26]Method Category=Detection based, Evaluation Protocol=Person-Centric (PC), PCK Threshold=0.22017.10 | 96.8 | 89 | 82.7 | 79.1 | 90.9 | 86 | 82.5 | — | — | — | — | — | — | 86.7 | — | — | — | |
| Lifshitz et al.publication=ECCV'162018.08 | 96.8 | 89 | 82.7 | 79.1 | 90.9 | 86 | 82.5 | 86.7 | — | — | — | — | — | — | — | — | — | |
| Lifshitz et al., arXiv'162016.09 | 96.8 | 89 | 82.7 | 79.1 | 90.9 | 86 | 82.5 | 86.7 | — | — | — | — | — | — | — | — | — | |
| Knowledge-Guided Deep Fractal Neural Networks2017.05 | 96.4 | — | — | — | — | — | — | 92.3 | 98.6 | 95.8 | 93.6 | 90.7 | 84.2 | — | — | — | — | |
| Part heatmap regressionBackbone=Residual2016.09 | 96.3 | 92.2 | 88.2 | 85.2 | 92.2 | 91.5 | 88.6 | 90.7 | — | — | — | — | — | — | — | — | — | |
| 4HG+AWBackbone=4-stack Hourglass, Loss=Adaptive Wing Loss2019.04 | 96.3 | 88.7 | 81.1 | 78.2 | 88.3 | 88.1 | 86.4 | 85.9 | — | — | — | — | — | — | — | — | — | |
| Rafi et al.Metric=PCK@0.22017.05 | 95.8 | 86.2 | 79.3 | 75 | 86.6 | 83.8 | 79.8 | 83.8 | — | — | — | — | — | — | — | — | — | |
| Pishchulin et al.venue=CVPR'162017.05 | 95.8 | — | — | — | — | — | — | 84.3 | 97 | 88.8 | 82 | 82.4 | 71.8 | — | — | — | — | |
| Rafi et al. [34]Method Category=Detection based, Evaluation Protocol=Person-Centric (PC), PCK Threshold=0.22017.10 | 95.8 | 86.2 | 79.3 | 75 | 86.6 | 83.8 | 79.8 | — | — | — | — | — | — | 83.8 | — | — | — | |
| Rafi et al., BMVC'16# Param=56M, Deployment Cost=28G2018.11 | 95.8 | 86.2 | 79.3 | 75 | 86.6 | 83.8 | 79.8 | — | — | — | — | — | — | — | 83.8 | 56.9 | — | |
| Insafutdinov et al.venue=ECCV'162017.05 | 95.4 | — | — | — | — | — | — | 87.8 | 97 | 90.6 | 86.9 | 86.1 | 79.5 | — | — | — | — | |
| [2]2017.04 | 95.2 | 89 | 81.5 | 77 | 83.7 | 87 | 82.8 | 85.2 | — | — | — | — | — | — | — | — | — | |
| Belagiannis et al.2018.05 | 95.2 | 89 | 81.5 | 77 | 83.7 | 87 | 82.8 | 85.2 | — | — | — | — | — | — | — | — | — | |
| Belagiannis & ZissermanMetric=PCK@0.22017.05 | 95.2 | 89 | 81.5 | 77 | 83.7 | 87 | 82.8 | 85.2 | — | — | — | — | — | — | — | — | — | |
| Bulat & Tzimiropoulosvenue=ECCV'162017.05 | 95.2 | — | — | — | — | — | — | 88.9 | 97.7 | 92.4 | 89.3 | 86.7 | 79.7 | — | — | — | — | |
| Belag. and Ziss. [4]Method Category=Detection based, Evaluation Protocol=Person-Centric (PC), PCK Threshold=0.22017.10 | 95.2 | 89 | 81.5 | 77 | 83.7 | 87 | 82.8 | — | — | — | — | — | — | 85.2 | — | — | — | |
| Belagiannis et al.publication=FG'172018.08 | 95.2 | 89 | 81.5 | 77 | 83.7 | 87 | 82.8 | 85.2 | — | — | — | — | — | — | — | — | — | |
| Wei et al.venue=CVPR'162017.05 | 95 | — | — | — | — | — | — | 88.3 | 98 | 82.2 | 89.1 | 85.8 | 77.9 | — | — | — | — | |
| Lifshitz et al.venue=ECCV'162017.05 | 94.8 | — | — | — | — | — | — | 84.3 | 97.3 | 88.8 | 84.4 | 80.6 | 71.4 | — | — | — | — | |
| Part heatmap regressionBackbone=VGG2016.09 | 94.8 | 86.6 | 79.5 | 73.5 | 88.1 | 83.2 | 78.5 | 83.5 | — | — | — | — | — | — | — | — | — | |
| DeepCut2019.04 | 94.6 | 86.8 | 79.9 | 75.4 | 83.5 | 82.8 | 77.9 | 83 | — | — | — | — | — | — | — | — | — | |
| 4HG+MSEBackbone=4-stack Hourglass, Loss=MSE2019.04 | 94.3 | 85.9 | 78.2 | 72 | 84.8 | 83.1 | 80.6 | 81.8 | — | — | — | — | — | — | — | — | — | |
| Rafi et al.venue=BMVC'162017.05 | 93.3 | — | — | — | — | — | — | 81.2 | 97.6 | 87.3 | 80.2 | 76.8 | 66.2 | — | — | — | — | |
| Fan et al.Metric=PCK@0.22017.05 | 92.4 | 75.2 | 65.3 | 64 | 76.7 | 68.3 | 70.4 | 73 | — | — | — | — | — | — | — | — | — | |
| Fan et al. [15]Method Category=Detection based, Evaluation Protocol=Person-Centric (PC), PCK Threshold=0.22017.10 | 92.4 | 75.2 | 65.3 | 64 | 75.7 | 68.3 | 70.4 | — | — | — | — | — | — | 73 | — | — | — | |
| Fan et al., CVPR'152016.09 | 92.4 | 75.2 | 65.3 | 64 | 75.7 | 68.3 | 70.4 | 73 | — | — | — | — | — | — | — | — | — | |
| Fan et al., CVPR'152018.11 | 92.4 | 75.2 | 65.3 | 64 | 75.7 | 68.3 | 70.4 | — | — | — | — | — | — | — | 73 | 43.2 | — | |
| Chen & YuilleMetric=PCK@0.22017.05 | 91.8 | 78.2 | 71.8 | 65.5 | 73.3 | 70.2 | 63.4 | 73.4 | — | — | — | — | — | — | — | — | — | |
| Chen and Yuille [7]Method Category=Detection based, Evaluation Protocol=Person-Centric (PC), PCK Threshold=0.22017.10 | 91.8 | 78.2 | 71.8 | 65.5 | 73.3 | 70.2 | 63.4 | — | — | — | — | — | — | 73.4 | — | — | — | |
| Chen & Yuille, NIPS'142016.09 | 91.8 | 78.2 | 71.8 | 65.5 | 73.3 | 70.2 | 63.4 | 73.4 | — | — | — | — | — | — | — | — | — | |
| Chen & Yuille, NIPS'142018.11 | 91.8 | 78.2 | 71.8 | 65.5 | 73.3 | 70.2 | 63.4 | — | — | — | — | — | — | — | 73.4 | 40.1 | — | |
| Yang et al.Metric=PCK@0.22017.05 | 90.6 | 78.1 | 73.8 | 68.8 | 74.8 | 69.9 | 58.9 | 73.6 | — | — | — | — | — | — | — | — | — | |
| Tompson et al.Metric=PCK@0.22017.05 | 90.6 | 79.2 | 67.9 | 63.4 | 69.5 | 71 | 64.2 | 72.3 | — | — | — | — | — | — | — | — | — | |
| Tompson et al. [41]Method Category=Detection based, Evaluation Protocol=Person-Centric (PC), PCK Threshold=0.22017.10 | 90.6 | 79.2 | 67.9 | 63.4 | 69.5 | 71 | 64.2 | — | — | — | — | — | — | 72.3 | — | — | — | |
| Yang et al. [44]Method Category=Detection based, Evaluation Protocol=Person-Centric (PC), PCK Threshold=0.22017.10 | 90.6 | 78.1 | 73.8 | 68.8 | 74.8 | 69.9 | 58.9 | — | — | — | — | — | — | 73.6 | — | — | — | |
| Yang et al., CVPR'162016.09 | 90.6 | 78.1 | 73.8 | 68.8 | 74.8 | 69.9 | 58.9 | 73.6 | — | — | — | — | — | — | — | — | — | |
| Tompson et al., NIPS'142016.09 | 90.6 | 79.2 | 67.9 | 63.4 | 69.5 | 71 | 64.2 | 72.3 | — | — | — | — | — | — | — | — | — | |
| Tompson et al., NIPS' 142018.11 | 90.6 | 79.2 | 67.9 | 63.4 | 69.5 | 71 | 64.2 | — | — | — | — | — | — | — | 72.3 | 47.3 | — | |
| Yang et al., CVPR'162018.11 | 90.6 | 78.1 | 73.8 | 68.8 | 74.8 | 69.9 | 58.9 | — | — | — | — | — | — | — | 73.6 | 39.3 | — | |
| Carreira et al. [6]Method Category=Regression based, Evaluation Protocol=Person-Centric (PC), PCK Threshold=0.22017.10 | 90.5 | 81.8 | 65.8 | 59.8 | 81.6 | 70.6 | 62 | — | — | — | — | — | — | 73.1 | — | — | — | |
| Carreira et al., CVPR'162016.09 | 90.5 | 81.8 | 65.8 | 59.8 | 81.6 | 70.6 | 62 | 73.1 | — | — | — | — | — | — | — | — | — | |
| Carreira et al., CVPR'162018.11 | 90.5 | 81.8 | 65.8 | 59.8 | 81.6 | 70.6 | 62 | — | — | — | — | — | — | — | 73.1 | 41.5 | — | |
| Belagiannis & Zissermanvenue=FG'172017.05 | 89.4 | — | — | — | — | — | — | 82.1 | 96 | 86.7 | 82.2 | 79.4 | 69.4 | — | — | — | — | |
| Yu et al.Metric=PCK@0.22017.05 | 87.2 | 88.2 | 82.4 | 76.3 | 91.4 | 85.8 | 78.7 | 84.3 | — | — | — | — | — | — | — | — | — | |
| Pishchulin et al. ICCV'13Metric=PCK@0.22017.05 | 87.2 | 56.7 | 46.7 | 38 | 61 | 57.5 | 52.7 | 57.1 | — | — | — | — | — | — | — | — | — | |
| Pishchulin et al. [32]Method Category=Detection based, Evaluation Protocol=Person-Centric (PC), PCK Threshold=0.22017.10 | 87.2 | 56.7 | 46.7 | 38 | 61 | 57.5 | 52.7 | — | — | — | — | — | — | 57.1 | — | — | — | |
| Yu et al. [46]Method Category=Detection based, Evaluation Protocol=Person-Centric (PC), PCK Threshold=0.22017.10 | 87.2 | 88.2 | 82.4 | 76.3 | 91.4 | 85.8 | 78.7 | — | — | — | — | — | — | 84.3 | — | — | — | |
| Yu et al., ECCV'162018.11 | 87.2 | 88.2 | 82.4 | 76.3 | 91.4 | 85.8 | 78.7 | — | — | — | — | — | — | — | 84.3 | 55.2 | — | |
| Fan et al.venue=CVPR'152017.05 | 86.6 | — | — | — | — | — | — | 70.1 | 95.4 | 77.7 | 69.8 | 62.8 | 49.1 | — | — | — | — | |
| Chen & Yuillevenue=NIPS'142017.05 | 85.6 | — | — | — | — | — | — | 73.6 | 96 | 77.2 | 72.2 | 69.7 | 58.1 | — | — | — | — | |
| Pishchulin et al. (2013)venue=ICCV'132017.05 | 85.1 | — | — | — | — | — | — | 58 | 88.7 | 63.6 | 58.4 | 46 | 35.2 | — | — | — | — | |
| Wang & LiMetric=PCK@0.22017.05 | 84.7 | 57.1 | 43.7 | 36.7 | 56.7 | 52.4 | 50.8 | 54.6 | — | — | — | — | — | — | — | — | — | |
| Yang et al.venue=CVPR'162017.05 | 83.9 | — | — | — | — | — | — | 74.8 | 95.6 | 78.5 | 71.8 | 72.2 | 61.8 | — | — | — | — | |
| Tompson et al.venue=NIPS'142017.05 | 83.7 | — | — | — | — | — | — | 66.6 | 90.3 | 70.4 | 61.1 | 63 | 51.2 | — | — | — | — |