CBCL-PR: A Cognitively Inspired Model for Class-Incremental Learning in Robotics

For most real-world applications, robots need to adapt and learn continually with limited data in their environments. In this paper, we consider the problem of Few-Shot class Incremental Learning (FSIL), which an AI agent is required incrementally from a few samples without forgetting it has previously learned. To solve problem, present novel framework inspired by theories concept learning hippocampus neocortex. Our represents object classes form sets clusters stores them memory. The replays generated old classes, avoid when new classes. approach evaluated on two classification datasets resulting state-of-the-art (SOTA) performance for class-incremental FSIL. We also evaluate our FSIL robot demonstrating that can classify large set household objects human assistance.