Composed Image Retrieval on CIRR 1.0 (val)
35Recall@1CoLLM
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| CoLLMVision Encoder=BLIP-L/16, Evaluation Protocol=Zero-shot, Incorporates LLM=true2025.03 | 35 | 78.6 | 94.2 | |
| Slerp-TATVision Encoder=BLIP-L/16, Evaluation Protocol=Zero-shot2025.03 | 34 | 72.7 | 88.9 | |
| Slerp-TATVision Encoder=OpenAI CLIP-L/14, Evaluation Protocol=Zero-shot2025.03 | 30.9 | 70.9 | 89.2 | |
| CoLLMVision Encoder=OpenAI CLIP-L/14, Evaluation Protocol=Zero-shot, Incorporates LLM=true2025.03 | 29.7 | 72.8 | 91.5 | |
| CoLLMVision Encoder=OpenAI CLIP-B/32, Evaluation Protocol=Zero-shot, Incorporates LLM=true2025.03 | 28.6 | 71.8 | 92.7 | |
| Slerp-TATVision Encoder=OpenAI CLIP-B/32, Evaluation Protocol=Zero-shot2025.03 | 28.2 | 68.8 | 88.5 | |
| MCLVision Encoder=OpenAI CLIP-L/14, Evaluation Protocol=Zero-shot, Incorporates LLM=true2025.03 | 25.9 | 69.4 | 91 | |
| ContextI2WVision Encoder=OpenAI CLIP-L/14, Evaluation Protocol=Zero-shot2025.03 | 25.7 | 68.6 | 89.6 | |
| LinCIRVision Encoder=OpenAI CLIP-L/14, Evaluation Protocol=Zero-shot, Reproduced Results=true2025.03 | 24.6 | 66.9 | 88.8 | |
| CIREVLVision Encoder=OpenAI CLIP-L/14, Evaluation Protocol=Zero-shot, Incorporates LLM=true2025.03 | 24.6 | 64.9 | 86.3 | |
| SEARLEVision Encoder=OpenAI CLIP-L/14, Evaluation Protocol=Zero-shot2025.03 | 24.2 | 66.3 | 88.8 | |
| SEARLEVision Encoder=OpenAI CLIP-B/32, Evaluation Protocol=Zero-shot2025.03 | 24 | 66.8 | 89.8 | |
| CIREVLVision Encoder=OpenAI CLIP-B/32, Evaluation Protocol=Zero-shot, Incorporates LLM=true2025.03 | 23.9 | 66 | 87 | |
| Pic2WorldVision Encoder=OpenAI CLIP-L/14, Evaluation Protocol=Zero-shot2025.03 | 23.9 | 65.3 | 87.8 | |
| PALAVRAVision Encoder=OpenAI CLIP-B/32, Evaluation Protocol=Zero-shot2025.03 | 16.6 | 58.5 | 84 |