List your company

Image Classification on DTD (Generic Accuracy)

43.99AccuracyCLIP ViT-B/16
21.858827.604433.3539.0956May 21, 2026

Evaluation results

MethodLinks
CLIP ViT-B/16Alignment train data=WIT400M, Zero-shot=true2026.0543.99
BEiT-B/16Alignment train data=IN21K repr. + 1 img-capt. pair per class, Zero-shot=true2026.0538.94
BEiT-B/16Alignment train data=IN21K repr., Zero-shot=true2026.0535.37
BEiT-B/16Alignment train data=IN1K repr., Zero-shot=true2026.0522.71