…We create it from MOVi dataset for amodal segmentation. The virtual camera is set to go around the scene, capturing about 24 consecutive frames.
1 PAPER • NO BENCHMARKS YET