Image Classification on TAG
58AccuracyOmniVaT
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| OmniVaTVenue=Ours, Backbone=ViT-B/16, Pre-trained=CLIP [31], Source Domain=TQ-GelMini2026.01 | 58 | — | 60 | |
| OmniVaTBackbone=ViT-B/16, Pre-trained=CLIP [31], Source Domain=TQ-T3D2026.01 | 56.5 | 59.1 | — | |
| LDCBackbone=ViT-B/16, Pre-trained=CLIP [31], Source Domain=TQ-T3D2026.01 | 56.1 | 43.2 | — | |
| SPGBackbone=ViT-B/16, Pre-trained=CLIP [31], Source Domain=TQ-T3D2026.01 | 54.3 | 53.8 | — | |
| SPGVenue=ECCV-24, Backbone=ViT-B/16, Pre-trained=CLIP [31], Source Domain=TQ-GelMini2026.01 | 53.7 | — | 50 | |
| CLIPBackbone=ViT-B/16, Pre-trained=CLIP [31], Source Domain=TQ-T3D2026.01 | 53 | 40.4 | — | |
| CLIPVenue=ICML-21, Backbone=ViT-B/16, Pre-trained=CLIP [31], Source Domain=TQ-GelMini2026.01 | 53 | — | 40.4 | |
| PromptStylerBackbone=ViT-B/16, Pre-trained=CLIP [31], Source Domain=TQ-T3D2026.01 | 52.9 | 48.8 | — | |
| LDCVenue=CVPR-25, Backbone=ViT-B/16, Pre-trained=CLIP [31], Source Domain=TQ-GelMini2026.01 | 52.9 | — | 48.5 | |
| MMRLBackbone=ViT-B/16, Pre-trained=CLIP [31], Source Domain=TQ-T3D2026.01 | 38.3 | 33.6 | — | |
| PromptStylerVenue=ICCV-23, Backbone=ViT-B/16, Pre-trained=CLIP [31], Source Domain=TQ-GelMini2026.01 | 35.8 | — | 33.2 | |
| MMRLVenue=CVPR-25, Backbone=ViT-B/16, Pre-trained=CLIP [31], Source Domain=TQ-GelMini2026.01 | 32.1 | — | 22.7 |