Human Part Segmentation On Cihp
Metrics
Mean IoU
Results
Performance results of various models on this benchmark
| Paper Title | ||
|---|---|---|
| Hulk(Finetune, ViT-L) | 72.68 | Hulk: A Universal Knowledge Translator for Human-Centric Tasks |
| Hulk(Finetune, ViT-B) | 71.26 | Hulk: A Universal Knowledge Translator for Human-Centric Tasks |
| UniHCP (finetune) | 69.8 | UniHCP: A Unified Model for Human-Centric Perceptions |
| ResNet101 | 67.47 | Self-Correction for Human Parsing |
| Parsing R-CNN + ResNext101 | 61.1 | Parsing R-CNN for Instance-Level Human Analysis |
| PGN + ResNet101 | 55.8 | Instance-level Human Parsing via Part Grouping Network |
0 of 6 row(s) selected.