ViT$^3$: Unlocking Test-Time Training in Vision
Open paper
Benchmark
Model
Rank
Results
semantic-segmentation-on-ade20k
H-ViT³-B + UperNet
#102
Validation mIoU: 51.7
image-generation-on-imagenet-256x256
DiT³-B/2
#200
FID: 39.31
image-classification-on-imagenet
H-ViT³-B
#240
Top 1 Accuracy: 85.5%
Rank counts only results with a code link.