GtA-SFDA Source-Only (DeepLabv2-ResNet101)

Reported on 2 benchmarks across 2 tasks · 1 paper · 2 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Methodology1 result

Domain AdaptationonGTA5-to-Cityscapes
mIoU· 2021-08-25
43.5
best: 74.4 (CLUDA+HRDA)
SOTA
Generalize then Adapt: Source-Free Domain Adaptive Semantic Segmentation arXiv:2108.11249

Computer Vision1 result

Domain GeneralizationonGTA5-to-Cityscapes
mIoU· 2021-08-25
43.5
best: 68.88 (tqdm (EVA02-CLIP-L))
SOTA
Generalize then Adapt: Source-Free Domain Adaptive Semantic Segmentation arXiv:2108.11249