arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2512.08673cs.CV
收藏

双分支中心-周围对比:重新思考3D点云中的对比学习

Dual-Branch Center-Surrounding Contrast: Rethinking Contrastive Learning for 3D Point Clouds

University of Science and Technology of China(中国科学技术大学);Shanghai Jiao Tong University(上海交通大学);Chinese University of Hong Kong(香港中文大学)

Shaofeng Zhang, Xuanqi Chen, Xiangdong Zhang, Sitong Wu, Junchi Yan

AI总结:

本文提出双分支中心-周围对比框架,通过双分支输入和补丁级对比损失,提升3D点云对比学习性能,达到SOTA效果。

摘要:

现有的基于3D点云的自监督学习(SSL)方法大多由基于掩码自动编码器(MAE)的生成方法主导。然而,这些生成方法已被证明在有效捕捉高层次判别特征方面存在困难,导致在线性探测和其他下游任务上的性能较差。相比之下,对比方法在图像数据中表现出色,在判别特征表示和泛化能力方面表现优异。尽管如此,3D数据中的对比学习(CL)仍然稀缺。此外,简单地将为2D数据设计的对比方法应用于3D数据,无法有效学习3D局部细节。为了解决这些挑战,我们提出了一种新的双分支中心-周围对比(CSCon)框架。具体来说,我们分别对中心和周围部分应用掩码,构建具有中心偏向和周围偏向表示的双分支输入,以更好地捕捉丰富的几何信息。同时,我们引入了基于补丁的对比损失,进一步增强高层信息和局部敏感性。在FULL和ALL协议下,CSCon的性能与生成方法相当;在MLP-LINEAR、MLP-3和ONLY-NEW协议下,我们的方法取得了最先进的结果,甚至超过了跨模态方法。特别是,在MLP-LINEAR协议下,我们的方法在ScanObjectNN的三个变体上分别优于基线(Point-MAE)7.9%、6.7%和10.3%。代码将公开发布。

英文摘要:

Most existing self-supervised learning (SSL) approaches for 3D point clouds are dominated by generative methods based on Masked Autoencoders (MAE). However, these generative methods have been proven to struggle to capture high-level discriminative features effectively, leading to poor performance on linear probing and other downstream tasks. In contrast, contrastive methods excel in discriminative feature representation and generalization ability on image data. Despite this, contrastive learning (CL) in 3D data remains scarce. Besides, simply applying CL methods designed for 2D data to 3D fails to effectively learn 3D local details. To address these challenges, we propose a novel Dual-Branch \textbf{C}enter-\textbf{S}urrounding \textbf{Con}trast (CSCon) framework. Specifically, we apply masking to the center and surrounding parts separately, constructing dual-branch inputs with center-biased and surrounding-biased representations to better capture rich geometric information. Meanwhile, we introduce a patch-level contrastive loss to further enhance both high-level information and local sensitivity. Under the FULL and ALL protocols, CSCon achieves performance comparable to generative methods; under the MLP-LINEAR, MLP-3, and ONLY-NEW protocols, our method attains state-of-the-art results, even surpassing cross-modal approaches. In particular, under the MLP-LINEAR protocol, our method outperforms the baseline (Point-MAE) by \textbf{7.9\%}, \textbf{6.7\%}, and \textbf{10.3\%} on the three variants of ScanObjectNN, respectively. The code will be made publicly available.

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。

↑