Improving Semantic Segmentation through

Spatio-Temporal Consistency Learned from Videos

Ankita Pasad, Ariel Gordon, Tsung-Yi Lin, Anelia Angelova