LiDAR semantic segmentation (LSS) for autonomous driving has been a growing field of interest in recent years. Datasets and methods have appeared and expanded very quickly, but methods have not been updated to exploit this new data availability and rely on the same classical datasets. Different ways of performing LSS training and inference can be divided into several subfields, which include the following: domain generalization, source-to-source segmentation, and pretraining. In this work, we aim to improve results in all of these subfields with the novel approach of multisource training. Multisource training relies on the availability of various datasets at training time. To overcome the common obstacles in multisource training, we introduce the coarse labels and call the newly created multisource dataset COLA. We propose three applications of this new dataset that display systematic improvement over single-source strategies: COLA-DG for domain generalization (+10% ), COLA-S2S for source-to-source segmentation (+5.3% ), and COLA-PT for pretraining (+12% ). We demonstrate that multisource approaches bring systematic improvement over single-source approaches.

Using deep learning, 3D autonomous driving semantic segmentation has become a well-studied subject, with methods that can reach very high performance. Nonetheless, because of the limited size of the training datasets, these models cannot see every type of object and scene found in real-world applications. The ability to be reliable in these various unknown environments is called domain generalization.Despite its importance, domain generalization is relatively unexplored in the case of 3D autonomous driving semantic segmentation. To fill this gap, this paper presents the first benchmark for this application by testing state-of-the-art methods and discussing the difficulty of tackling Laser Imaging Detection and Ranging (LiDAR) domain shifts.We also propose the first method designed to address this domain generalization, which we call 3DLabelProp. This method relies on leveraging the geometry and sequentiality of the LiDAR data to enhance its generalization performances by working on partially accumulated point clouds. It reaches a mean Intersection over Union (mIoU) of 50.4% on SemanticPOSS and of 55.2% on PandaSet solid-state LiDAR while being trained only on SemanticKITTI, making it the state-of-the-art method for generalization (+5% and +33% better, respectively, than the second best method).The code for this method is available on GitHub: https://github.com/JulesSanchez/3DLabelProp.

Transfer learning is a proven technique in 2D computer vision to leverage the large amount of data available and achieve high performance with datasets limited in size due to the cost of acquisition or annotation. In 3D, annotation is known to be a costly task; nevertheless, pre-training methods have only recently been investigated. Due to this cost, unsupervised pretraining has been heavily favored. In this work, we tackle the case of real-time 3D semantic segmentation of sparse autonomous driving LiDAR scans. Such datasets have been increasingly released, but each has a unique label set. We propose here an intermediate-level label set called coarse labels, which can easily be used on any existing and future autonomous driving datasets, thus allowing all the data available to be leveraged at once without any additional manual labeling. This way, we have access to a larger dataset, alongside a simple task of semantic segmentation. With it, we introduce a new pretraining task: coarse label pre-training, also called COLA. We thoroughly analyze the impact of COLA on various datasets and architectures and show that it yields a noticeable performance improvement, especially when only a small dataset is available for the finetuning task.