Text-to-Image Retrieval on COCO (test)
4,610R@1GENIUS with Re-ranking
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| GENIUS with Re-rankingTraining Data=COCO, Zero-shot=false, Re-ranking=true2025.03 | 4,610 | 7,400 | — | — | 8,270 | — | |
| GENIUS with Re-rankingTraining Data=M-BEIR, Zero-shot=true, Re-ranking=true2025.03 | 4,260 | 6,730 | — | — | 7,890 | — | |
| GENIUSTraining Data=COCO, Zero-shot=false2025.03 | 4,120 | 6,780 | — | — | 7,780 | — | |
| GENIUSTraining Data=M-BEIR, Zero-shot=true2025.03 | 4,000 | 6,550 | — | — | 7,680 | — | |
| IRGenTraining Data=COCO2025.03 | 2,960 | 5,070 | — | — | 5,630 | — | |
| GRACE (Structured ID)Training Data=COCO2025.03 | 1,670 | 3,920 | — | — | 5,030 | — | |
| GRACE (Semantic ID)Training Data=COCO2025.03 | 1,330 | 3,040 | — | — | 3,590 | — | |
| CoMPGFLOPs=29.24±0.322026.04 | 57.32 | 81.44 | — | — | — | — | |
| MADTPGFLOPs=30.04±0.442026.04 | 56.28 | 80.74 | — | — | — | — | |
| SigLIPZero-shot=true2025.09 | 47.8 | — | — | — | — | — | |
| DreamLIPZero-shot=true2025.09 | 47.7 | — | — | — | — | — | |
| Long-CLIPBackbone=ViT-L/14, Zero-shot=true2025.03 | 46.96 | 71.89 | 90.25 | 95.36 | — | — | |
| CE-CLIPZero-shot=true2025.09 | 46.9 | — | — | — | — | — | |
| GOAL fine-tuned with DOCCIBackbone=ViT-L/14, Zero-shot=true2025.03 | 46.29 | 70.85 | 89.43 | 95.2 | — | — | |
| GOAL fine-tuned with DCIBackbone=ViT-L/14, Zero-shot=true2025.03 | 45.54 | 70.22 | 89.09 | 94.9 | — | — | |
| NegCLIPZero-shot=true2025.09 | 44.8 | — | — | — | — | — | |
| Long-CLIPBackbone=ViT-B/16, Zero-shot=true2025.03 | 40.83 | 66.36 | 87.42 | 93.97 | — | — | |
| LongCLIPZero-shot=true2025.09 | 40.6 | — | — | — | — | — | |
| GOAL fine-tuned with DCIBackbone=ViT-B/16, Zero-shot=true2025.03 | 39.08 | 65.32 | 86.93 | 93.66 | — | — | |
| LSSZero-shot=true2025.09 | 38.9 | — | — | — | — | — | |
| GOAL fine-tuned with DOCCIBackbone=ViT-B/16, Zero-shot=true2025.03 | 38.86 | 64.36 | 86.22 | 93.28 | — | — | |
| DACLLMZero-shot=true2025.09 | 37.7 | — | — | — | — | — | |
| CLIPBackbone=ViT-L/14, Zero-shot=true2025.03 | 37.29 | 61.82 | 84.19 | 91.83 | — | — | |
| CLIPBackbone=ViT-B/16, Zero-shot=true2025.03 | 33.95 | 59.46 | 82.95 | 91.06 | — | — | |
| CLIPZero-shot=true2025.09 | 30.2 | — | — | — | — | — | |
| DCIP1Zero-shot=true2025.09 | 21.4 | — | — | — | — | — | |
| GRACE (String ID)Training Data=COCO2025.03 | 12 | 37 | — | — | 88 | — | |
| GRACE (Numeric ID)Training Data=COCO2025.03 | 3 | 14 | — | — | 28 | — | |
| CovMatch# Pairs=100, Compression rate=0.8‰2026.06 | — | — | — | — | — | 6.5 | |
| CovMatch# Pairs=200, Compression rate=1.7‰2026.06 | — | — | — | — | — | 7.4 | |
| CovMatch# Pairs=500, Compression rate=4.4‰2026.06 | — | — | — | — | — | 9.9 | |
| Forgetting# Pairs=100, Compression rate=0.8‰2026.06 | — | — | — | — | — | 1.4 | |
| Forgetting# Pairs=200, Compression rate=1.7‰2026.06 | — | — | — | — | — | 2.8 | |
| Forgetting# Pairs=500, Compression rate=4.4‰2026.06 | — | — | — | — | — | 4.9 | |
| Herding# Pairs=100, Compression rate=0.8‰2026.06 | — | — | — | — | — | 3 | |
| Herding# Pairs=200, Compression rate=1.7‰2026.06 | — | — | — | — | — | 4.8 | |
| Herding# Pairs=500, Compression rate=4.4‰2026.06 | — | — | — | — | — | 8.6 | |
| K-Center# Pairs=100, Compression rate=0.8‰2026.06 | — | — | — | — | — | 3.3 | |
| K-Center# Pairs=200, Compression rate=1.7‰2026.06 | — | — | — | — | — | 5.1 | |
| K-Center# Pairs=500, Compression rate=4.4‰2026.06 | — | — | — | — | — | 8.9 | |
| LoRS# Pairs=100, Compression rate=0.8‰2026.06 | — | — | — | — | — | 7 | |
| LoRS# Pairs=200, Compression rate=1.7‰2026.06 | — | — | — | — | — | 9.1 | |
| LoRS# Pairs=500, Compression rate=4.4‰2026.06 | — | — | — | — | — | 9.7 | |
| MTT-VL# Pairs=100, Compression rate=0.8‰2026.06 | — | — | — | — | — | 5.4 | |
| MTT-VL# Pairs=200, Compression rate=1.7‰2026.06 | — | — | — | — | — | 6.8 | |
| MTT-VL# Pairs=500, Compression rate=4.4‰2026.06 | — | — | — | — | — | 9.1 | |
| RAHA# Pairs=100, Compression rate=0.8‰2026.06 | — | — | — | — | — | 6.6 | |
| RAHA# Pairs=200, Compression rate=1.7‰2026.06 | — | — | — | — | — | 9.3 | |
| RAHA# Pairs=500, Compression rate=4.4‰2026.06 | — | — | — | — | — | 12.6 | |
| Random# Pairs=100, Compression rate=0.8‰2026.06 | — | — | — | — | — | 2.9 | |
| Random# Pairs=200, Compression rate=1.7‰2026.06 | — | — | — | — | — | 4.7 | |
| Random# Pairs=500, Compression rate=4.4‰2026.06 | — | — | — | — | — | 8.6 | |
| TESLA_BCE# Pairs=100, Compression rate=0.8‰2026.06 | — | — | — | — | — | 1 | |
| TESLA_BCE# Pairs=200, Compression rate=1.7‰2026.06 | — | — | — | — | — | 0.3 | |
| TESLA_BCE# Pairs=500, Compression rate=4.4‰2026.06 | — | — | — | — | — | 3.7 |