List your company

Image-to-Text Retrieval on FLICKR30K (R@1, R@5, R@10)

97.3R@1BLIP
-1.81223.91949.6575.381Dec 17, 2025Jan 6, 2026Jan 26, 2026Feb 15, 2026Mar 7, 2026Mar 27, 2026Apr 17, 2026

Evaluation results

MethodLinks
BLIP# Tunable=223M, Fine-tuning strategy=full fine-tuning2025.1297.399.9100
DoRA (r=32)# Tunable=10.8M, Fine-tuning strategy=frozen backbone, rank=322025.1296.799.8100
UniAdapter (r=512)# Tunable=19.5M, Fine-tuning strategy=frozen backbone, rank=5122025.1296.799.7100
Aurora (r=64)# Tunable=0.6M, Fine-tuning strategy=frozen backbone, rank=642025.1296.799.8100
Null-LoRA# Tunable=6.0M, Fine-tuning strategy=frozen backbone2025.1296.6100100
LoRA (r=32)# Tunable=10.6M, Fine-tuning strategy=frozen backbone, rank=322025.1296.399.799.8
ALBEF# Tunable=210M, Fine-tuning strategy=full fine-tuning2025.1295.999.8100
ALIGN# Tunable=820M, Fine-tuning strategy=full fine-tuning2025.1295.399.8100
VILLA# Tunable=330M, Fine-tuning strategy=full fine-tuning2025.1287.997.598.8
UNITER# Tunable=330M, Fine-tuning strategy=full fine-tuning2025.1287.39899.2
UniConBackbone=CLIP ViT-B/32, Train time=0.76 s2026.0428.4—63.6
SGD–CLIPBackbone=CLIP ViT-B/32, Train time=45.3 s2026.0423.1—59.5
UniConBackbone=RN-50 + SBERT, Train time=0.81 s2026.0413.4—46.4
SGD–CLIPBackbone=RN-18 + SBERT, Train time=45.6 s2026.044.3—22.1
SGD–CLIPBackbone=RN-50 + SBERT, Train time=45.0 s2026.044.3—22.1
UniConBackbone=RN-18 + SBERT, Train time=1.7 s2026.042—14.5