CCL：面向激光雷达地点识别的持续对比学习
CCL: Continual Contrastive Learning for LiDAR Place Recognition
AI总结：
针对激光雷达地点识别的深度学习方法存在泛化差和灾难性遗忘问题，提出构建对比特征池、结合分布知识蒸馏的持续对比学习方法CCL，在多数据集上验证其性能优于现有最优持续学习方法。
摘要：
地点识别是机器人与自动驾驶应用中回环检测与全局定位的一项核心且具挑战性的任务。得益于深度学习技术的最新进展，激光雷达地点识别（LiDAR place recognition, LPR）的性能已得到大幅提升。然而，当前基于深度学习的方法存在两大核心问题：泛化能力差与灾难性遗忘。\n本文提出一种名为CCL的持续对比学习方法，旨在解决灾难性遗忘问题，并普遍提升LPR方法的鲁棒性。我们的CCL构建了一个对比特征池，利用对比损失训练出更具可迁移性的地点表征。当迁移到新环境时，CCL会持续回顾对比记忆库，并采用基于分布的知识蒸馏，在持续学习从新数据中识别新地点的同时，维持对过往数据的检索能力。\n我们使用三种不同的LPR方法，在Oxford、MulRan和PNV数据集上对所提方法进行了全面评估。实验结果表明，CCL在不同环境下持续提升了不同方法的性能，表现优于当前最优的持续学习方法。本方法的实现代码已发布于https://github.com/cloudcjf/CCL。
英文摘要：
Place recognition is an essential and challenging task in loop closing and global localization for robotics and autonomous driving applications. Benefiting from the recent advances in deep learning techniques, the performance of LiDAR place recognition (LPR) has been greatly improved. However, current deep learning-based methods suffer from two major problems: poor generalization ability and catastrophic forgetting. In this paper, we propose a continual contrastive learning method, named CCL, to tackle the catastrophic forgetting problem and generally improve the robustness of LPR approaches. Our CCL constructs a contrastive feature pool and utilizes contrastive loss to train more transferable representations of places. When transferred into new environments, our CCL continuously reviews the contrastive memory bank and applies a distribution-based knowledge distillation to maintain the retrieval ability of the past data while continually learning to recognize new places from the new data. We thoroughly evaluate our approach on Oxford, MulRan, and PNV datasets using three different LPR methods. The experimental results show that our CCL consistently improves the performance of different methods in different environments outperforming the state-of-the-art continual learning method. The implementation of our method has been released at https://github.com/cloudcjf/CCL.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
