DTL: Disentangled Transfer Learning for Visual Recognition

Open paper
Benchmark Model Rank Results
monocular-depth-estimation-on-clevr-distanceDTL+ ViT-B/16 (VTAB-1k)–Accuracy: 65.1
object-counting-on-clevr-countDTL ViT-B/16 (VTAB-1k)–Accuracy: 82.8

Rank counts only results with a code link.