Segmenting Transparent Object in the Wild with Transformer

Open paper

Benchmark	Model	Rank	Results
semantic-segmentation-on-trans10k	Trans2Seg	#3	mIoU: 72.15%GFLOPs: 49.03