Mix3D: Out-of-Context Data Augmentation for 3D Scenes

Open paper

Benchmark	Model	Rank	Results
semantic-segmentation-on-scannet	Mix3D	#24	val mIoU: 73.6test mIoU: 78.1