Human Part Parsing on PASCAL-Person-Part (test)
74.96mIoUSST (Transfer-ATR&CIHP)
Evaluation results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| SST (Transfer-ATR&CIHP)Backbone=ResNet-1012023.04 | 74.96 | — | — | — | — | — | — | — | — | |
| SST (Transfer-CIHP)Backbone=ResNet-1012023.04 | 74.52 | — | — | — | — | — | — | — | — | |
| PLIPBackbone=Swin-B2023.05 | 73.93 | — | — | — | — | — | — | — | — | |
| SST (Transfer-ATR&CIHP)Backbone=ResNet-502023.04 | 73.87 | — | — | — | — | — | — | — | — | |
| PLIPBackbone=RN1522023.05 | 73.51 | — | — | — | — | — | — | — | — | |
| SST (Transfer-CIHP)Backbone=ResNet-502023.04 | 73.49 | — | — | — | — | — | — | — | — | |
| SST (Universal)Backbone=ResNet-1012023.04 | 73.13 | — | — | — | — | — | — | — | — | |
| PLIPBackbone=RN1012023.05 | 72.63 | — | — | — | — | — | — | — | — | |
| SST (Universal)Backbone=ResNet-502023.04 | 72.25 | — | — | — | — | — | — | — | — | |
| PLIPBackbone=RN502023.05 | 72.14 | — | — | — | — | — | — | — | — | |
| NPPNet2023.04 | 71.73 | — | — | — | — | — | — | — | — | |
| Grapy-MLBackbone=Xception2023.04 | 71.65 | — | — | — | — | — | — | — | — | |
| SNTBackbone=ResNet-1012023.04 | 71.59 | — | — | — | — | — | — | — | — | |
| SCHPBackbone=ResNet-1012023.04 | 71.56 | — | — | — | — | — | — | — | — | |
| SCHPBackbone=RN1012023.05 | 71.46 | — | — | — | — | — | — | — | — | |
| Multi-datasetsBackbone=ResNet-1012023.04 | 71.41 | — | — | — | — | — | — | — | — | |
| DPCBackbone=Xception2023.04 | 71.34 | — | — | — | — | — | — | — | — | |
| GraphonomySource Dataset=CIHP2021.01 | 71.14 | — | — | — | — | — | — | — | — | |
| GraphonomyBackbone=Xception2023.04 | 71.14 | — | — | — | — | — | — | — | — | |
| GraphonomySource Dataset=Universal Human Parsing2021.01 | 70.99 | — | — | — | — | — | — | — | — | |
| LearningBackbone=ResNet-1012023.04 | 70.76 | — | — | — | — | — | — | — | — | |
| CNIFBackbone=RN1012023.05 | 70.76 | — | — | — | — | — | — | — | — | |
| Multi-datasetsBackbone=ResNet-502023.04 | 70.4 | — | — | — | — | — | — | — | — | |
| GPMBackbone=Xception2023.04 | 69.5 | — | — | — | — | — | — | — | — | |
| RefineNet2021.01 | 68.6 | — | — | — | — | — | — | — | — | |
| Bilinski et al.2021.01 | 68.6 | — | — | — | — | — | — | — | — | |
| PGN2021.01 | 68.4 | — | — | — | — | — | — | — | — | |
| PGNBackbone=Xception2023.04 | 68.4 | — | — | — | — | — | — | — | — | |
| Single-datasetBackbone=ResNet-1012023.04 | 68.32 | — | — | — | — | — | — | — | — | |
| Multi-task Learning2021.01 | 68.13 | — | — | — | — | — | — | — | — | |
| DeepLab v3+2021.01 | 67.84 | — | — | — | — | — | — | — | — | |
| Deeplab V3+Backbone=ResNet-1012023.04 | 67.84 | — | — | — | — | — | — | — | — | |
| Pose-Guided Knowledge TransferBackbone=ResNet-101, Multi-scale testing=true2018.05 | 67.6 | 87.15 | 72.28 | 57.07 | 56.21 | 52.43 | 50.36 | 97.72 | — | |
| Fang et al.2021.01 | 67.6 | — | — | — | — | — | — | — | — | |
| Deeplab V3+Backbone=Xception2023.04 | 67.6 | — | — | — | — | — | — | — | — | |
| Single-datasetBackbone=ResNet-502023.04 | 67.15 | — | — | — | — | — | — | — | — | |
| Li et al.2021.01 | 66.3 | — | — | — | — | — | — | — | — | |
| DeepLab v22021.01 | 64.94 | — | — | — | — | — | — | — | — | |
| JointBackbone=ResNet-101, Multi-scale testing=true2018.05 | 64.39 | 85.5 | 67.87 | 54.72 | 54.3 | 48.25 | 44.76 | 95.32 | — | |
| Our modelBackbone=ResNet-101, Pose Integration=true2017.08 | 64.39 | 85.5 | 67.87 | 54.72 | 54.3 | 48.25 | 44.76 | 95.32 | — | |
| Pose-Guided Knowledge TransferBackbone=ResNet-101, Multi-scale testing=false2018.05 | 64.28 | 84.83 | 68.64 | 53.11 | 53.01 | 48.4 | 46.76 | 95.22 | — | |
| Structure-evolving LSTMMulti-scale testing=false2018.05 | 63.57 | 82.89 | 67.15 | 51.42 | 48.72 | 51.72 | 45.91 | 97.18 | — | |
| Structure-evolving LSTM2021.01 | 63.57 | — | — | — | — | — | — | — | — | |
| Our modelBackbone=ResNet-101, Pose Integration=false2017.08 | 62.66 | 84.95 | 67.21 | 52.81 | 51.37 | 46.27 | 41.03 | 94.96 | — | |
| Pose-Guided Knowledge TransferBackbone=VGG-16, Multi-scale testing=false2018.05 | 62.6 | 84.06 | 67.03 | 51.66 | 50.15 | 45.33 | 44.26 | 95.73 | — | |
| Graph LSTMMulti-scale testing=false2018.05 | 60.16 | 82.69 | 62.68 | 46.88 | 47.71 | 45.66 | 40.93 | 94.59 | — | |
| Attention + MMANmulti-scale=true, plug-and-play-module=attention network2018.07 | 59.91 | 82.58 | 62.83 | 48.49 | 47.37 | 42.8 | 40.4 | 94.92 | — | |
| MMANBackbone=RN1012023.05 | 59.91 | — | — | — | — | — | — | — | — | |
| Attention + SSL2018.07 | 59.36 | 83.26 | 62.4 | 47.8 | 45.58 | 42.32 | 39.48 | 94.68 | — | |
| LIP2021.01 | 59.36 | — | — | — | — | — | — | — | — | |
| JPPNetBackbone=RN1012023.05 | 59.36 | — | — | — | — | — | — | — | — | |
| MMAN2018.07 | 58.45 | 82.46 | 61.41 | 46.05 | 45.17 | 40.93 | 38.83 | 94.3 | — | |
| JointBackbone=VGG-16, Multi-scale testing=true2018.05 | 58.06 | 80.21 | 61.36 | 47.53 | 43.94 | 41.77 | 38 | 93.64 | — | |
| Our modelBackbone=VGG-16, Pose Integration=true2017.08 | 58.06 | 80.21 | 61.36 | 47.53 | 43.94 | 41.77 | 38 | 93.64 | — | |
| LG-LSTM2018.07 | 57.97 | 82.72 | 60.99 | 45.4 | 47.76 | 42.33 | 37.96 | 88.63 | — | |
| Macro AN2018.07 | 57.58 | 82.01 | 61.19 | 45.24 | 44.3 | 39.73 | 36.75 | 93.89 | — | |
| HAZNMulti-scale testing=false2018.05 | 57.54 | 80.79 | 60.76 | 45.65 | 43.11 | 41.21 | 37.74 | 93.78 | — | |
| HAZN2017.08 | 57.54 | 80.76 | 60.5 | 45.65 | 43.11 | 41.21 | 37.74 | 93.78 | — | |
| Micro AN2018.07 | 57.23 | 82.44 | 61.35 | 44.79 | 43.68 | 38.41 | 36.05 | 93.93 | — | |
| Do-Deeplab-ASPP2018.07 | 56.79 | 81.82 | 59.53 | 44.8 | 42.79 | 38.32 | 36.38 | 93.91 | — | |
| Our modelBackbone=VGG-16, Pose Integration=false2017.08 | 56.5 | 79.83 | 59.72 | 43.84 | 40.84 | 40.49 | 37.23 | 93.55 | — | |
| AttentionMulti-scale testing=false2018.05 | 56.39 | 81.47 | 59.06 | 44.15 | 42.5 | 38.28 | 35.62 | 93.65 | — | |
| Attention2018.07 | 56.39 | 81.47 | 59.06 | 44.15 | 42.5 | 38.28 | 35.62 | 93.65 | — | |
| Attention2017.08 | 56.39 | 81.47 | 59.06 | 44.15 | 42.5 | 38.28 | 35.62 | 93.65 | — | |
| AttentionBackbone=VGG162023.05 | 56.39 | — | — | — | — | — | — | — | — | |
| HAZN2018.07 | 56.11 | 80.79 | 59.11 | 43.05 | 42.76 | 38.99 | 34.46 | 93.59 | — | |
| Deeplab-ASPP2018.07 | 55.19 | 81.33 | 60.06 | 41.16 | 40.95 | 37.49 | 32.56 | 92.81 | — | |
| DeepLab-LargeFOV-CRFMulti-scale testing=false2018.05 | 52.95 | 80.13 | 55.56 | 36.43 | 38.72 | 35.5 | 30.82 | 93.52 | — | |
| Attention2018.08 | — | 81.47 | 59.06 | 44.15 | 42.5 | 38.28 | 35.62 | 93.65 | 56.39 | |
| Attention2017.03 | — | 81.47 | 59.06 | 44.15 | 42.5 | 38.28 | 35.62 | 93.65 | 56.39 | |
| Attention2018.04 | — | 81.47 | 59.06 | 44.15 | 42.5 | 38.28 | 35.62 | 93.65 | 56.39 | |
| Attention + SSLStrategy=Self-supervised structure-sensitive learning2017.03 | — | 83.26 | 62.4 | 47.8 | 45.58 | 42.32 | 39.48 | 94.68 | 59.36 | |
| DeepLab v22018.08 | — | — | — | — | — | — | — | — | 64.94 | |
| DeepLab-LargeFOV2015.11 | — | 78.09 | 54.02 | 37.29 | 36.85 | 33.73 | 29.61 | 92.85 | 51.78 | |
| DeepLab-LargeFOV2017.03 | — | 78.09 | 54.02 | 37.29 | 36.85 | 33.73 | 29.61 | 92.85 | 51.78 | |
| DeepLab-LargeFOV2018.04 | — | 78.09 | 54.02 | 37.29 | 36.85 | 33.73 | 29.61 | 92.85 | 51.78 | |
| DeepLab-LargeFOV-CRFpost-processing=Conditional Random Field (CRF)2015.11 | — | 80.13 | 55.56 | 36.43 | 38.72 | 35.5 | 30.82 | 93.52 | 52.95 | |
| Graph LSTM2018.08 | — | 82.69 | 62.68 | 46.88 | 47.71 | 45.66 | 40.93 | 94.59 | 60.16 | |
| HAZN2018.08 | — | 80.79 | 59.11 | 43.05 | 42.76 | 38.99 | 34.46 | 93.59 | 56.11 | |
| HAZN2017.03 | — | 80.79 | 59.11 | 43.05 | 42.76 | 38.99 | 34.46 | 93.59 | 56.11 | |
| HAZN2018.04 | — | 80.79 | 59.11 | 43.05 | 42.76 | 38.99 | 34.46 | 93.59 | 56.11 | |
| HAZN (full model)configuration=full model2015.11 | — | 80.76 | 60.5 | 45.65 | 43.11 | 41.21 | 37.74 | 93.78 | 57.54 | |
| HAZN (no object scale)configuration=no object scale2015.11 | — | 80.25 | 57.2 | 42.24 | 42.02 | 36.4 | 31.96 | 93.42 | 54.78 | |
| HAZN (no part scale)configuration=no part scale2015.11 | — | 79.83 | 59.72 | 43.84 | 40.84 | 40.49 | 37.23 | 93.55 | 56.5 | |
| Holistic2018.08 | — | — | — | — | — | — | — | — | 66.3 | |
| LG-LSTM2018.08 | — | 82.72 | 60.99 | 45.4 | 47.76 | 42.33 | 37.96 | 88.63 | 57.97 | |
| LG-LSTM2017.03 | — | 82.72 | 60.99 | 45.4 | 47.76 | 42.33 | 37.96 | 88.63 | 57.97 | |
| LG-LSTM2018.04 | — | 82.72 | 60.99 | 45.4 | 47.76 | 42.33 | 37.96 | 88.63 | 57.97 | |
| LIP2018.08 | — | 83.26 | 62.4 | 47.8 | 45.58 | 42.32 | 39.48 | 94.68 | 59.36 | |
| Multi-Scale Attention2015.11 | — | 81.47 | 59.06 | 44.15 | 42.5 | 38.28 | 35.62 | 93.65 | 56.39 | |
| Multi-Scale Averagingscales=0.5, 1.0, 1.52015.11 | — | 79.89 | 57.4 | 40.57 | 41.14 | 37.66 | 34.31 | 93.43 | 54.91 | |
| PGNrefinement=true, full_model=true2018.08 | — | 90.89 | 75.12 | 55.83 | 64.61 | 55.42 | 41.57 | 95.33 | 68.4 | |
| PGN (segmentation)component=segmentation branch only2018.08 | — | 89.98 | 73.7 | 54.75 | 60.26 | 50.58 | 39.16 | 95.09 | 66.22 | |
| PGN (w/o refinement)refinement=false2018.08 | — | 90.11 | 72.93 | 54.01 | 59.47 | 54.57 | 42.03 | 95.12 | 66.91 | |
| SS-JPPNetbackbone=Attention [8], inference mode=multi-scale and flipped2018.04 | — | 83.26 | 62.4 | 47.8 | 45.58 | 42.32 | 39.48 | 94.68 | 59.36 | |
| Structure-evolving LSTM2018.08 | — | 82.89 | 67.15 | 51.42 | 48.72 | 51.72 | 45.91 | 97.18 | 63.57 |