DatUS^2: Data-driven Unsupervised Semantic Segmentation with Pre-trained Self-supervised Vision Transformer
Open paper
Benchmark
Model
Rank
Results
unsupervised-semantic-segmentation-on-suim
DatUS (ViT-B/8) + OC
#1
Pixel Accuracy: 69.98
mIoU: 34.02
unsupervised-semantic-segmentation-on-suim
DatUS (ViT-B/8)
#3
Pixel Accuracy: 64.67
mIoU: 28.48
Rank counts only results with a code link.