| CoLLM (finetuned - BLIP-L/16) |
- R@1
- 45.8
- R@10
- 84.7
- R@50
- 95.8
|
CoLLM: A Large Language Model for Composed Image Retrieval
2025-03-25 |
코드 1 |
| CoVR-BLIP-2 |
- R@5
- 73.61
- R@1
- 43.74
- R@10
- 83.95
- R@50
- 96.1
|
CoVR-2: Automatic Data Construction for Composed Video Retrieval
2023-08-28 |
코드 1 |
| ImageScope (CLIP-ViT-L/14) |
- R@5
- 67.54
- R@1
- 39.37
- R@10
- 78.05
- R@50
- 92.94
|
ImageScope: Unifying Language-Guided Image Retrieval via Large Multimodal Model Collective Reasoning
2025-03-13 |
코드 1 |
| IP-CIR + LDRE (CLIP G/14) |
- R@5
- 70.07
- R@1
- 39.25
- R@10
- 80
- R@50
- 94.89
|
Imagine and Seek: Improving Composed Image Retrieval with an Imagined Proxy
2024-11-24 |
연결 없음 |
| SEIZE (CLIP G/14) |
- R@5
- 69.42
- R@1
- 38.87
- R@10
- 79.42
|
Semantic Editing Increment Benefits Zero-Shot Composed Image Retrieval
2024-10-28 |
코드 1 · 코드 2 |
| TransAgg (Laion-CIR-Combined) |
- R@5
- 68.88
- R@1
- 37.87
- R@50
- 93.86
|
Zero-shot Composed Text-Image Retrieval
2023-06-12 |
코드 1 |
| OSrCIR (CLIP G/14) |
- R@5
- 67.25
- R@1
- 37.26
- R@10
- 77.33
|
Reason-before-Retrieve: One-Stage Reflective Chain-of-Thoughts for Training-Free Zero-Shot Composed Image Retrieval
2024-12-15 |
코드 1 |
| SCOT (WACV 2025) |
- R@5
- 64.34
- R@1
- 36.82
- R@10
- 74.48
- R@50
- 93.42
|
SCOT: Self-Supervised Contrastive Pretraining For Zero-Shot Compositional Retrieval
2025-01-12 |
연결 없음 |
| CoLLM (Pretrained - BLIP-L/16) |
- R@1
- 35.00
- R@10
- 78.6
- R@50
- 94.2
|
CoLLM: A Large Language Model for Composed Image Retrieval
2025-03-25 |
코드 1 |
| CIReVL (CLIP G/14) |
- R@5
- 64.29
- R@1
- 34.65
|
Vision-by-Language for Training-Free Compositional Image Retrieval
2023-10-13 |
코드 1 |
| MagicLens (CoCa L) |
- R@5
- 67.0
- R@1
- 33.3
- R@10
- 77.9
- R@50
- 94.4
|
MagicLens: Self-Supervised Image Retrieval with Open-Ended Instructions
2024-03-28 |
코드 1 |
| WeiMoCIR (CLIP G/14) |
- R@5
- 60.41
- R@1
- 31.04
|
Training-free Zero-shot Composed Image Retrieval via Weighted Modality Fusion and Similarity
2024-09-07 |
코드 1 |
| WeiMoCIR (CLIP L/14) |
- R@5
- 60.87
- R@1
- 30.94
|
Training-free Zero-shot Composed Image Retrieval via Weighted Modality Fusion and Similarity
2024-09-07 |
코드 1 |
| MagicLens (CLIP L) |
- R@5
- 61.7
- R@1
- 30.1
- R@10
- 74.4
- R@50
- 92.6
|
MagicLens: Self-Supervised Image Retrieval with Open-Ended Instructions
2024-03-28 |
코드 1 |
| IP-CIR + LDRE (CLIP L/14) |
- R@5
- 58.82
- R@1
- 29.76
- R@10
- 71.21
- R@50
- 90.41
|
Imagine and Seek: Improving Composed Image Retrieval with an Imagined Proxy
2024-11-24 |
연결 없음 |
| CoLLM (Pretrained - CLIP-L/14) |
- R@1
- 29.7
- R@10
- 72.8
- R@50
- 91.5
|
CoLLM: A Large Language Model for Composed Image Retrieval
2025-03-25 |
코드 1 |
| OSrCIR (CLIP L/14) |
- R@5
- 57.68
- R@1
- 29.45
- R@10
- 69.86
|
Reason-before-Retrieve: One-Stage Reflective Chain-of-Thoughts for Training-Free Zero-Shot Composed Image Retrieval
2024-12-15 |
코드 1 |
| WeiMoCIR (CLIP H/14) |
- R@5
- 59.76
- R@1
- 29.11
|
Training-free Zero-shot Composed Image Retrieval via Weighted Modality Fusion and Similarity
2024-09-07 |
코드 1 |
| WeiMoCIR (CLIP B/32) |
- R@5
- 57.69
- R@1
- 26.31
|
Training-free Zero-shot Composed Image Retrieval via Weighted Modality Fusion and Similarity
2024-09-07 |
코드 1 |
| OSrCIR (CLIP B/32) |
- R@5
- 54.54
- R@1
- 25.42
- R@10
- 68.19
|
Reason-before-Retrieve: One-Stage Reflective Chain-of-Thoughts for Training-Free Zero-Shot Composed Image Retrieval
2024-12-15 |
코드 1 |
| CIReVL (CLIP L/14) |
- R@5
- 52.31
- R@1
- 24.55
|
Vision-by-Language for Training-Free Compositional Image Retrieval
2023-10-13 |
코드 1 |
| CIReVL (CLIP B/32) |
- R@5
- 52.51
- R@1
- 23.94
|
Vision-by-Language for Training-Free Compositional Image Retrieval
2023-10-13 |
코드 1 |
| RTD + LinCIR (CLIP G/14) |
- R@5
- 67.47
|
An Efficient Post-hoc Framework for Reducing Task Discrepancy of Text Encoders for Composed Image Retrieval
2024-06-13 |
코드 1 |
| LDRE (CLIP G/14) |
- R@5
- 66.39
|
LDRE: LLM-based Divergent Reasoning and Ensemble for Zero-Shot Composed Image Retrieval
2024-07-11 |
코드 1 · 코드 2 |
| SPN4CIR (SPN-CC) |
- R@5
- 65.42
- Rsubset@1
- 64.87
|
Improving Composed Image Retrieval via Contrastive Learning with Scaling Positives and Negatives
2024-04-17 |
코드 1 |
| LinCIR (CLIP G/14) |
- R@5
- 64.72
|
Language-only Efficient Training of Zero-shot Composed Image Retrieval
2023-12-04 |
코드 1 |
| MagicLens (CoCa B) |
- R@5
- 64.0
|
MagicLens: Self-Supervised Image Retrieval with Open-Ended Instructions
2024-03-28 |
코드 1 |
| MTCIR (BLIP B/16) |
- R@5
- 58.87
|
Pretrain like Your Inference: Masked Tuning Improves Zero-Shot Composed Image Retrieval
2023-11-13 |
코드 1 |
| MagicLens (CLIP B) |
- R@5
- 58.0
|
MagicLens: Self-Supervised Image Retrieval with Open-Ended Instructions
2024-03-28 |
코드 1 |
| CompoDiff (CLIP G/14) |
- R@5
- 57.61
|
CompoDiff: Versatile Composed Image Retrieval With Latent Diffusion
2023-03-21 |
코드 1 |