xMUDA: Cross-Modal Unsupervised Domain Adaptation for 3D Semantic Segmentation | IEEE Conference Publication | IEEE Xplore