List your company
ResearchFormerly WizWand

Text-to-Image Retrieval on MSCOCO (test)

84.6R@5X-VLMclip
2.824824.054945.28566.5151Oct 14, 2022Feb 16, 2023Jun 21, 2023Oct 25, 2023Feb 27, 2024Jul 1, 2024Nov 4, 2024
Updated 8mo ago

Evaluation results

MethodLinks
X-VLMclip2022.1084.661.590.8
EfficientVLM2022.1084.460.690.5
X-VLMclipperformance_target=98%2022.1082.960.389
X-VLMsmall2022.1081.656.188.7
OSCARB2022.1080.85488.5
OSCARBperformance_target=98%2022.1079.252.986.7
FFF-ViT-B/32Pre-train dataset=Open70M, Zero-shot=true2024.0574.949.183.2
MiniVLM2022.1074.14584
DistillVLM2022.1073.743.983.3
ViLT2022.1072.942.783.1
FFF-ViT-B/32Pre-train dataset=Open30M, Zero-shot=true2024.0572.646.481.6
FFF-ViT-B/32Pre-train dataset=YFCC-15M, Zero-shot=true2024.0570.944.380.1
MLCDPre-training Data=LAION-400M, Backbone=ViT-L/14, Evaluation Mode=Zero-shot2024.0769.644.579.9
FLIPPre-training Data=LAION-400M, Backbone=ViT-L/14, Evaluation Mode=Zero-shot2024.0769.244.278.4
OpenCLIPPre-training Data=LAION-400M, Backbone=ViT-L/14, Evaluation Mode=Zero-shot2024.0766.641.376.1
CLIPPre-training Data=WIT-400M, Backbone=ViT-L/14, Evaluation Mode=Zero-shot2024.0760.735.870.7
TripletCLIPTraining Dataset=CC12M, Evaluation Protocol=Zero-shot, Backbone=ViT-B/322024.1128.5——
NegCLIPTraining Dataset=CC12M, Evaluation Protocol=Zero-shot, Backbone=ViT-B/322024.1123.11——
NegCLIP++Training Dataset=CC12M, Evaluation Protocol=Zero-shot, Backbone=ViT-B/322024.1122.69——
LaCLIPTraining Dataset=CC12M, Evaluation Protocol=Zero-shot, Backbone=ViT-B/322024.1119.78——
TripletCLIPTraining Dataset=CC3M, Evaluation Protocol=Zero-shot, Backbone=ViT-B/322024.1111.28——
LaCLIP + HNTraining Dataset=CC3M, Evaluation Protocol=Zero-shot, Backbone=ViT-B/322024.118.64——
NegCLIPTraining Dataset=CC3M, Evaluation Protocol=Zero-shot, Backbone=ViT-B/322024.116.61——
NegCLIP++Training Dataset=CC3M, Evaluation Protocol=Zero-shot, Backbone=ViT-B/322024.116.19——
LaCLIPTraining Dataset=CC3M, Evaluation Protocol=Zero-shot, Backbone=ViT-B/322024.115.97——