SqueezeSegV2 swMATH ID: 36660 Software Authors: Bichen Wu, Xuanyu Zhou, Sicheng Zhao, Xiangyu Yue, Kurt Keutzer Description: SqueezeSegV2: Improved Model Structure and Unsupervised Domain Adaptation for Road-Object Segmentation from a LiDAR Point Cloud. Earlier work demonstrates the promise of deep-learning-based approaches for point cloud segmentation; however, these approaches need to be improved to be practically useful. To this end, we introduce a new model SqueezeSegV2 that is more robust to dropout noise in LiDAR point clouds. With improved model structure, training loss, batch normalization and additional input channel, SqueezeSegV2 achieves significant accuracy improvement when trained on real data. Training models for point cloud segmentation requires large amounts of labeled point-cloud data, which is expensive to obtain. To sidestep the cost of collection and annotation, simulators such as GTA-V can be used to create unlimited amounts of labeled, synthetic data. However, due to domain shift, models trained on synthetic data often do not generalize well to the real world. We address this problem with a domain-adaptation training pipeline consisting of three major components: 1) learned intensity rendering, 2) geodesic correlation alignment, and 3) progressive domain calibration. When trained on real data, our new model exhibits segmentation accuracy improvements of 6.0-8.6 Homepage: https://arxiv.org/abs/1809.08495 Source Code: https://github.com/xuanyuzhou98/SqueezeSegV2 Keywords: Computer Vision; Pattern Recognition; arXiv_cs.CV Related Software: PointSIFT; VoxNet; OctNet; SnapNet; PointNet; InteriorNet; YOLO; SYNTHIA Dataset; SPLATNet; RGCNN; VoxSegNet; ScanNet; ScanComplete; SceneNN; DeepLab; Paris-Lille-3D; ImageNet; PASCAL VOC; Semantic3D.net; ShapeNet Cited in: 1 Publication Cited by 4 Authors 1 Gao, Jiangjiang 1 Lan, Jinhui 1 Li, Feifan 1 Wang, Bingxu Cited in 1 Serial 1 Machine Learning Cited in 1 Field 1 Computer science (68-XX) Citations by Year