SELMA: SEmantic Large-scale Multimodal Acquisitions in Variable Weather, Daytime and Viewpoints
This addresses the data scarcity problem for autonomous driving researchers by providing a free, diverse synthetic dataset, though it is incremental as it builds on existing simulators like CARLA.
The authors tackled the lack of large-scale, multi-sensor synthetic datasets for semantic segmentation in autonomous driving by introducing SELMA, a dataset with over 20M samples across 27 weather and daytime conditions, which enables efficient training and achieves strong results on real-world data.
Accurate scene understanding from multiple sensors mounted on cars is a key requirement for autonomous driving systems. Nowadays, this task is mainly performed through data-hungry deep learning techniques that need very large amounts of data to be trained. Due to the high cost of performing segmentation labeling, many synthetic datasets have been proposed. However, most of them miss the multi-sensor nature of the data, and do not capture the significant changes introduced by the variation of daytime and weather conditions. To fill these gaps, we introduce SELMA, a novel synthetic dataset for semantic segmentation that contains more than 30K unique waypoints acquired from 24 different sensors including RGB, depth, semantic cameras and LiDARs, in 27 different atmospheric and daytime conditions, for a total of more than 20M samples. SELMA is based on CARLA, an open-source simulator for generating synthetic data in autonomous driving scenarios, that we modified to increase the variability and the diversity in the scenes and class sets, and to align it with other benchmark datasets. As shown by the experimental evaluation, SELMA allows the efficient training of standard and multi-modal deep learning architectures, and achieves remarkable results on real-world data. SELMA is free and publicly available, thus supporting open science and research.