Open-world Semantic Segmentation for LIDAR Point Clouds
This addresses the robustness issue for autonomous driving systems by enabling them to handle unseen objects and adapt over time, though it is incremental as it builds on existing open-set and incremental learning methods.
The paper tackles the problem of LIDAR semantic segmentation being closed-set and static, which limits real-world applications like autonomous driving, by proposing an open-world approach that identifies both old and novel classes and incorporates them incrementally without forgetting; the REAL framework achieves state-of-the-art performance on SemanticKITTI and nuScenes datasets and significantly reduces catastrophic forgetting.
Current methods for LIDAR semantic segmentation are not robust enough for real-world applications, e.g., autonomous driving, since it is closed-set and static. The closed-set assumption makes the network only able to output labels of trained classes, even for objects never seen before, while a static network cannot update its knowledge base according to what it has seen. Therefore, in this work, we propose the open-world semantic segmentation task for LIDAR point clouds, which aims to 1) identify both old and novel classes using open-set semantic segmentation, and 2) gradually incorporate novel objects into the existing knowledge base using incremental learning without forgetting old classes. For this purpose, we propose a REdundAncy cLassifier (REAL) framework to provide a general architecture for both the open-set semantic segmentation and incremental learning problems. The experimental results show that REAL can simultaneously achieves state-of-the-art performance in the open-set semantic segmentation task on the SemanticKITTI and nuScenes datasets, and alleviate the catastrophic forgetting problem with a large margin during incremental learning.