MLAE: Masked LoRA Experts for Visual Parameter-Efficient Fine-Tuning
Open paper
Benchmark
Model
Rank
Results
monocular-depth-estimation-on-clevr-distance
MLAE ViT-B/16 (VTAB-1k)
–
Accuracy: 67.2
object-counting-on-clevr-count
MLAE ViT-B/16 (VTAB-1k)
–
Accuracy: 84.3
Rank counts only results with a code link.