PointJEM：通过联合熵最大化减少特征冗余的自监督点云理解
PointJEM: Self-supervised Point Cloud Understanding for Reducing Feature Redundancy via Joint Entropy Maximization
AI总结：
针对自监督点云表征中的特征冗余问题，提出 PointJEM，通过分块嵌入与联合熵最大化使特征两两独立，并在分类、分割等下游任务上取得竞争力性能。
摘要：
大多数基于深度学习的点云处理方法都是监督式的，需要大规模标注数据。然而，人工标注点云数据既费力又耗时。自监督表征学习能够通过从未标注数据集中学习鲁棒且具有泛化性的表征来解决上述问题。尽管如此，表征学习得到的嵌入特征通常包含冗余信息，而当前大多数方法通过线性相关约束来减少特征冗余。本文提出 PointJEM，这是一种应用于点云领域的自监督表征学习方法。PointJEM 包含一种嵌入方案和一个基于联合熵的损失函数。该嵌入方案将嵌入向量划分为不同部分，每个部分都能学习一种独特特征。为减少特征中的冗余信息，PointJEM 最大化不同部分之间的联合熵，从而使学习到的特征变量两两独立。为验证该方法的有效性，我们在多个数据集上进行了实验。结果表明，该方法能够显著减少线性相关之外的特征冗余。此外，PointJEM 在分类和分割等下游任务中取得了具有竞争力的性能。
英文摘要：
Most deep learning-based point cloud processing methods are supervised and require large scale of labeled data. However, manual labeling of point cloud data is laborious and time-consuming. Self-supervised representation learning can address the aforementioned issue by learning robust and generalized representations from unlabeled datasets. Nevertheless, the embedded features obtained by representation learning usually contain redundant information, and most current methods reduce feature redundancy by linear correlation constraints. In this paper, we propose PointJEM, a self-supervised representation learning method applied to the point cloud field. PointJEM comprises an embedding scheme and a loss function based on joint entropy. The embedding scheme divides the embedding vector into different parts, each part can learn a distinctive feature. To reduce redundant information in the features, PointJEM maximizes the joint entropy between the different parts, thereby rendering the learned feature variables pairwise independent. To validate the effectiveness of our method, we conducted experiments on multiple datasets. The results demonstrate that our method can significantly reduce feature redundancy beyond linear correlation. Furthermore, PointJEM achieves competitive performance in downstream tasks such as classification and segmentation.
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。
