Scene Sketch Semantic Segmentation on FrISS
82.24Accuracy@PLASA
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| LASASupervision/Evaluation Protocol=Weakly supervised, Backbone=ViT-B/162026.06 | 82.24 | 73.87 | 73.97 | |
| CLIPerSupervision/Evaluation Protocol=Fine-tuned, Backbone=ViT-B/162026.06 | 79.81 | 71.93 | 71.39 | |
| ProxyCLIPSupervision/Evaluation Protocol=Fine-tuned, Backbone=ViT-B/162026.06 | 76.01 | 68.55 | 65.49 | |
| CLIPerSupervision/Evaluation Protocol=Training free2026.06 | 75.87 | 70.14 | 67.82 | |
| ProxyCLIPSupervision/Evaluation Protocol=Training free2026.06 | 75.46 | 67 | 65 | |
| CorrCLIPSupervision/Evaluation Protocol=Fine-tuned, Backbone=ViT-B/162026.06 | 72.3 | 71.01 | 63.99 | |
| CorrCLIPSupervision/Evaluation Protocol=Training free2026.06 | 69.9 | 65.43 | 58.94 | |
| BourouisSupervision/Evaluation Protocol=Weakly supervised2026.06 | 67.34 | 59.92 | 58.23 | |
| CLIPSupervision/Evaluation Protocol=Fine-tuned, Backbone=ViT-B/162026.06 | 54.66 | 57.78 | 28.01 | |
| CLIPSupervision/Evaluation Protocol=Training free2026.06 | 42.57 | 43.19 | 21.03 | |
| SketchSegerSupervision/Evaluation Protocol=Fully supervised2026.06 | 38.87 | 34.78 | 17.35 |