Consolidator: Mergeable Adapter with Grouped Connections for Visual Adaptation
Open paper
Benchmark
Model
Rank
Results
monocular-depth-estimation-on-clevr-distance
Consolidator ViT-B/16 (VTAB-1k)
–
Accuracy: 68.2
object-counting-on-clevr-count
Consolidator ViT-B/16 (VTAB-1k)
–
Accuracy: 81.2
Rank counts only results with a code link.