SePiCo: Semantic-Guided Pixel Contrast for Domain Adaptive Semantic Segmentation

Open paper
Benchmark Model Rank Results
image-to-image-translation-on-synthia-to-cityscapesSePiCo#4mIoU (13 classes): 71.4
unsupervised-domain-adaptation-on-synthia-to-cityscapesSePiCo#6mIoU (13 classes): 71.4
unsupervised-domain-adaptation-on-synthia-to-cityscapesSePiCo (DeepLabv2 ResNet-101)#11mIoU (13 classes): 66.5
semantic-segmentation-on-dark-zurichSePiCo#7mIoU: 54.2
semantic-segmentation-on-dark-zurichSePiCo (DeepLab v2 ResNet-101)#10mIoU: 45.4
semantic-segmentation-on-synthia-to-cityscapesSePiCo#7Mean IoU: 64.3
image-to-image-translation-on-gtav-to-cityscapes-labelsSePiCo#8mIoU: 70.3
semantic-segmentation-on-gtav-to-cityscapes-labelsSePiCo#8mIoU: 70.3
synthetic-to-real-translation-on-synthia-to-cityscapesSePiCo#9MIoU (16 classes): 64.3MIoU (13 classes): 71.4
synthetic-to-real-translation-on-synthia-to-cityscapesSePiCo (ResNet-101)#16MIoU (16 classes): 58.1MIoU (13 classes): 66.5
domain-adaptation-on-gta5-to-cityscapesSePiCo#11mIoU: 70.3
domain-adaptation-on-synthia-to-cityscapesSePiCo#11mIoU: 64.3
domain-adaptation-on-synthia-to-cityscapesSePiCo (DeepLabv2-ResNet-101)#18mIoU: 58.1
synthetic-to-real-translation-on-gtav-to-cityscapes-labelsSePiCo#11mIoU: 70.3
synthetic-to-real-translation-on-gtav-to-cityscapes-labelsSePiCo - DeepLabv2#21mIoU: 61.0
unsupervised-domain-adaptation-on-gtav-to-cityscapes-labelsSePiCo#11mIoU: 70.3

Rank counts only results with a code link.