Image-to-Text Retrieval on Bookset zero-shot
34.4Recall@1 (zero-shot)HOMIE
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| HOMIEModel Category=MLLM-Based Models, Backbone=Qwen3-VL-8B, Evaluation Protocol=Zero-shot2025.02 | 34.4 | 61.4 | 73 | |
| HOMIEModel Category=MLLM-Based Models, Backbone=Qwen3-VL-2B, Evaluation Protocol=Zero-shot2025.02 | 32 | 58.5 | 69.4 | |
| Patho-CLIPModel Category=Pathology Dual-Encoder Models, Evaluation Protocol=Zero-shot2025.02 | 26.7 | 55.1 | 67.2 | |
| CONCHModel Category=Pathology Dual-Encoder Models, Evaluation Protocol=Zero-shot2025.02 | 25 | 50 | 62.2 | |
| MUSKModel Category=Pathology Dual-Encoder Models, Evaluation Protocol=Zero-shot2025.02 | 23.9 | 49 | 62.3 | |
| Patho-CLIP-LModel Category=Pathology Dual-Encoder Models, Evaluation Protocol=Zero-shot2025.02 | 22.5 | 50.6 | 62 | |
| Pathgen-CLIPModel Category=Pathology Dual-Encoder Models, Evaluation Protocol=Zero-shot2025.02 | 15 | 32.3 | 41.4 | |
| Pathgen-CLIP-LModel Category=Pathology Dual-Encoder Models, Evaluation Protocol=Zero-shot2025.02 | 12.8 | 28.7 | 37.2 | |
| BiomedCLIPModel Category=Pathology Dual-Encoder Models, Evaluation Protocol=Zero-shot2025.02 | 9.9 | 23.5 | 32 | |
| PathoR1-7BModel Category=MLLM-Based Models, Backbone=7B, Evaluation Protocol=Zero-shot2025.02 | 9.5 | 23.8 | 32.3 | |
| PathCLIPModel Category=Pathology Dual-Encoder Models, Evaluation Protocol=Zero-shot2025.02 | 9.3 | 22.7 | 30.7 | |
| QuiltNetModel Category=Pathology Dual-Encoder Models, Evaluation Protocol=Zero-shot2025.02 | 3.6 | 12.6 | 18.2 | |
| PLIPModel Category=Pathology Dual-Encoder Models, Evaluation Protocol=Zero-shot2025.02 | 3.3 | 9.2 | 14.2 | |
| Lingshu-7BModel Category=MLLM-Based Models, Backbone=7B, Evaluation Protocol=Zero-shot2025.02 | 2.9 | 8.6 | 13.2 | |
| GME-7BModel Category=MLLM-Based Models, Backbone=7B, Evaluation Protocol=Zero-shot2025.02 | 1.9 | 5.1 | 7.8 | |
| E5-V-7BModel Category=MLLM-Based Models, Backbone=7B, Evaluation Protocol=Zero-shot2025.02 | 1.7 | 5.5 | 8.6 | |
| Qwen2.5-VL-7BModel Category=MLLM-Based Models, Backbone=7B, Evaluation Protocol=Zero-shot2025.02 | 0.9 | 2.8 | 4.7 | |
| Qwen3-VL-8BModel Category=MLLM-Based Models, Backbone=8B, Evaluation Protocol=Zero-shot2025.02 | 0.9 | 3.1 | 4.3 | |
| HuatuoGPT-V-7BModel Category=MLLM-Based Models, Backbone=7B, Evaluation Protocol=Zero-shot2025.02 | 0.9 | 4.6 | 7 | |
| LamRA-7BModel Category=MLLM-Based Models, Backbone=7B, Evaluation Protocol=Zero-shot2025.02 | 0.6 | 2.1 | 3.4 | |
| Qwen3-VL-2BModel Category=MLLM-Based Models, Backbone=2B, Evaluation Protocol=Zero-shot2025.02 | 0.5 | 1.7 | 2.8 | |
| PubmedCLIPModel Category=Pathology Dual-Encoder Models, Evaluation Protocol=Zero-shot2025.02 | 0.2 | 1 | 1.6 |