List your company
ResearchFormerly WizWand

Image Classification on Entity-30 OOD

72.9AccuracyLP-FT + VRF
64.78866.8946971.106Nov 11, 2024
Updated 8mo ago

Evaluation results

MethodLinks
LP-FT + VRFBackbone=CLIP ViT-B/162024.1172.9
E2E-FT + VRFBackbone=CLIP ViT-B/162024.1172.7
LP-FT + WSEBackbone=CLIP ViT-B/162024.1172.1
E2E-FT + WSEBackbone=CLIP ViT-B/162024.1171.9
LP-FT + OSEBackbone=CLIP ViT-B/162024.1171.8
E2E-FT + OSEBackbone=CLIP ViT-B/162024.1171.5
LP-FT + VRFBackbone=CLIP ViT-B/322024.1170.1
Linear classifierBackbone=CLIP ViT-B/162024.1169.6
E2E-FT + VRFBackbone=CLIP ViT-B/322024.1169.5
E2E-FT + WSEBackbone=CLIP ViT-B/322024.1168.8
LP-FT + WSEBackbone=CLIP ViT-B/322024.1168.8
LP-FTBackbone=CLIP ViT-B/162024.1168.8
LP-FT + OSEBackbone=CLIP ViT-B/322024.1168.5
Zero-shotBackbone=CLIP ViT-B/162024.1168.2
E2E-FTBackbone=CLIP ViT-B/162024.1168.2
Linear classifierBackbone=CLIP ViT-B/322024.1168.1
LP-FTBackbone=CLIP ViT-B/322024.1167.7
Zero-shotBackbone=CLIP ViT-B/322024.1166.5
E2E-FT + OSEBackbone=CLIP ViT-B/322024.1166.4
E2E-FTBackbone=CLIP ViT-B/322024.1165.1