…It not only supports video panoptic segmentation (VPS) task, but also provides super-set annotations for video semantic segmentation (VSS) and video instance segmentation (VIS) tasks.
24 PAPERS • 1 BENCHMARK