Expectation-Maximization Attention Networks for Semantic Segmentation

Open paper

Benchmark	Model	Rank	Results
semantic-segmentation-on-bdd100k-val	EMANet	#8	mIoU: 61.4
semantic-segmentation-on-coco-stuff-test	EMANet	#12	mIoU: 39.9%
semantic-segmentation-on-pascal-context	EMANet	#38	mIoU: 53.1