1 paper · 1 filter
Di Lin, Jifeng Dai, Jiaya Jia +2
Large-scale data is of crucial importance for learning semantic segmentation models, but annotating per-pixel masks is a tedious and inefficient procedure. We note that for the top…