GtA-SFDA (DeepLabv2-ResNet101)

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

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

Methodology2 results

Domain AdaptationonSYNTHIA-to-Cityscapes
mIoU· 2021-08-25
60.1
best: 78.1 (HALO)
SOTA
Generalize then Adapt: Source-Free Domain Adaptive Semantic Segmentation arXiv:2108.11249
Domain AdaptationonGTA5 to Cityscapes
mIoU· 2021-08-25
53.4
best: 77.8 (HALO)
Generalize then Adapt: Source-Free Domain Adaptive Semantic Segmentation arXiv:2108.11249