CoDEPS: Online Continual Learning for Depth Estimation and Panoptic Segmentation
This addresses the need for robust, self-adapting perception systems in robots operating in changing conditions, representing an incremental advance in continual learning methods.
The paper tackles the problem of online continual learning for depth estimation and panoptic segmentation in robotics, introducing CoDEPS to adapt to new environments without forgetting previous ones, achieving state-of-the-art results on real-world datasets.
Operating a robot in the open world requires a high level of robustness with respect to previously unseen environments. Optimally, the robot is able to adapt by itself to new conditions without human supervision, e.g., automatically adjusting its perception system to changing lighting conditions. In this work, we address the task of continual learning for deep learning-based monocular depth estimation and panoptic segmentation in new environments in an online manner. We introduce CoDEPS to perform continual learning involving multiple real-world domains while mitigating catastrophic forgetting by leveraging experience replay. In particular, we propose a novel domain-mixing strategy to generate pseudo-labels to adapt panoptic segmentation. Furthermore, we explicitly address the limited storage capacity of robotic systems by leveraging sampling strategies for constructing a fixed-size replay buffer based on rare semantic class sampling and image diversity. We perform extensive evaluations of CoDEPS on various real-world datasets demonstrating that it successfully adapts to unseen environments without sacrificing performance on previous domains while achieving state-of-the-art results. The code of our work is publicly available at http://codeps.cs.uni-freiburg.de.