…Each frame has a semantic segmentation of the objects in the scene and information about the camera pose. It is composed by 415 sequences captured in 254 different spaces, in 41 different buildings.
114 PAPERS • NO BENCHMARKS YET