Refign (DAFormer)

Reported on 5 benchmarks across 3 tasks · 1 paper

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

Medical2 results

Semantic SegmentationonDark Zurich
mIoU· 2022-07-14
56.2
best: 63.9 (Refign (HRDA))
Refign: Align and Refine for Adaptation of Semantic Segmentation to Adverse Conditions arXiv:2207.06825
Semantic SegmentationonNighttime Driving
mIoU· 2022-07-14
56.8
best: 60.8 (TADP)
Refign: Align and Refine for Adaptation of Semantic Segmentation to Adverse Conditions arXiv:2207.06825

Audio2 results

10-shot image generationonDark Zurich
mIoU· 2022-07-14
56.2
best: 63.9 (Refign (HRDA))
Refign: Align and Refine for Adaptation of Semantic Segmentation to Adverse Conditions arXiv:2207.06825
10-shot image generationonNighttime Driving
mIoU· 2022-07-14
56.8
best: 60.8 (TADP)
Refign: Align and Refine for Adaptation of Semantic Segmentation to Adverse Conditions arXiv:2207.06825

Methodology1 result

Domain AdaptationonCityscapes to ACDC
mIoU· 2022-07-14
65.5
best: 78.8 (SoRA)
Refign: Align and Refine for Adaptation of Semantic Segmentation to Adverse Conditions arXiv:2207.06825