Visual Entity Recognition on OVEN
28.5HM (Unseen)WikiCLIP-L
Evaluation Results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| WikiCLIP-LCategory=Contrastive, Latency (ms)=14.49, TFLOPS=1.932026.03 | 28.5 | — | — | — | — | 35.5 | 31.6 | |
| WikiCLIP-SCategory=Contrastive, Latency (ms)=14.49, TFLOPS=1.932026.03 | 27 | — | — | — | — | 36.8 | 31.1 | |
| GiT-Large*Category=Generative, Extra Dataset=REW-47M [3], Latency (ms)=83.95, TFLOPS=3.062026.03 | 25.1 | — | — | — | — | 36 | 29.6 | |
| AutoVER 13BCategory=Generative, Latency (ms)=1569, TFLOPS=24.742026.03 | 24.5 | — | — | — | — | 63.6 | 35.6 | |
| AutoVER 7BCategory=Generative, Latency (ms)=993, TFLOPS=19.472026.03 | 21.7 | — | — | — | — | 61.5 | 32.1 | |
| GPT4VCategory=Zero Shot2026.03 | 19.3 | — | — | — | — | 29.8 | 23.4 | |
| GER-ALD*Category=Generative, Extra Dataset=Entity-WebLI [4], Latency (ms)=83.95, TFLOPS=3.062026.03 | 17.7 | — | — | — | — | 31.5 | 22.7 | |
| GiT-Large*Category=Generative, Extra Dataset=Entity-WebLI [4], Latency (ms)=83.95, TFLOPS=3.062026.03 | 16.4 | — | — | — | — | 25.9 | 20.1 | |
| GPT5-nanoCategory=Zero Shot2026.03 | 13 | — | — | — | — | 23.7 | 16.8 | |
| PaLI-17BCategory=Generative2026.03 | 12.4 | — | — | — | — | 30.6 | 17.6 | |
| CLIP2CLIPCategory=Contrastive, Latency (ms)=13.84, TFLOPS=0.082026.03 | 10.5 | — | — | — | — | 12.6 | 11.5 | |
| PaLI-3BCategory=Generative2026.03 | 6.6 | — | — | — | — | 21.6 | 10.1 | |
| CLIP ViTL14Category=Contrastive, Latency (ms)=11.69, TFLOPS=0.072026.03 | 5.4 | — | — | — | — | 5.3 | 5.4 | |
| CLIPFusionCategory=Contrastive, Latency (ms)=15.93, TFLOPS=0.082026.03 | 4.8 | — | — | — | — | 33.6 | 8.4 | |
| GiT-Large*Category=Generative, Extra Dataset=WebLI-100M [5], Latency (ms)=83.95, TFLOPS=3.062026.03 | 4.2 | — | — | — | — | 13.7 | 6.5 | |
| CLIP I-IRetrieval Mode=Image-to-Image2026.03 | — | 10.1 | 27.1 | 36 | 44.1 | — | — | |
| WikiCLIP-LScale=Large2026.03 | — | 31.6 | 53.3 | 60.5 | 67.4 | — | — | |
| WikiCLIP-SScale=Small2026.03 | — | 31.1 | 53.7 | 61.3 | 67.8 | — | — |