Recurrent Scene Parsing with Perspective Understanding in the Loop

Benchmark Model Rank Results
semantic-segmentation-on-cityscapesDepthSeg (ResNet-101)#56Mean IoU (class): 78.2%
semantic-segmentation-on-nyu-depth-v2RecurrentSceneParsing#66Mean IoU: 44.5%
semantic-segmentation-on-sun-rgbdDPLNet#32Mean IoU: 45.1%