Context-Aware Mixup for Domain Adaptive Semantic Segmentation

Open paper
Benchmark Model Rank Results
image-to-image-translation-on-gtav-toCAMix (w DAFormer)#6mIoU: 70.0
image-to-image-translation-on-gtav-toCAMix (w Deeplabv2 ResNet 101)#14mIoU: 55.2
image-to-image-translation-on-synthia-toCAMix (w DAFormer)#5mIoU (13 classes): 69.2
image-to-image-translation-on-synthia-toCAMix (w Deeplabv2 ResNet 101)#11mIoU (13 classes): 59.7
synthetic-to-real-translation-on-gtav-toCAMix (w DAFormer)#9mIoU: 70.0
synthetic-to-real-translation-on-gtav-toCAMix (w Deeplabv2 ResNet101)#26mIoU: 55.2
synthetic-to-real-translation-on-synthia-to-1CAMix (w DAFormer)#29MIoU (13 classes): 69.2
synthetic-to-real-translation-on-synthia-to-1CAMix (ResNet 101)#30MIoU (13 classes): 59.7
unsupervised-domain-adaptation-on-gtav-toCAMix (w DAFormer)#8mIoU: 70.0
unsupervised-domain-adaptation-on-gtav-toCAMix (w Deeplabv2 ResNet 101)#14mIoU: 55.2

Rank counts only results with a code link.