SemiGMMPoint: Semi-supervised point cloud segmentation based on Gaussian mixture models

• Published in: Pattern recognition Vol. 158; p. 111045
• Main Authors: Zhuang, Xianwei, Wang, Hualiang, He, Xiaoxuan, Fu, Siming, Hu, Haoji
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.02.2025
• Subjects: 3D point cloud segmentation; Contrastive learning; Distribution similarity minimization; Gaussian mixture models; Generative classifier; Semi-supervised learning; Gaussian mixture models; 3D point cloud segmentation; Generative classifier; Semi-supervised learning; Distribution similarity minimization; Contrastive learning
• DOI: 10.1016/j.patcog.2024.111045
• ISSN: 0031-3203

Existing semi-supervised point cloud segmentation methods emphasize on discriminative learning, which overlooks the underlying class-conditional distributions and distribution similarities. In this paper, we propose SemiGMMPoint, the first generative framework for semi-supervised 3D point cloud segmentation in real-world and large-scale settings. Specifically, we propose a point dense generative classifier based on Gaussian mixture models (GMMs) to explicitly estimate class-conditional distributions. On top of it, we incorporate a novel similarity-minimization algorithm into the Expectation–Maximization (EM) based GMM parameter estimation, which minimizes the inter-class distribution similarity in the representation space. Moreover, we utilize the well-calibrated posterior to develop a modified point contrastive loss to mitigate sampling bias in semi-supervised settings. Extensive experiments show that SemiGMMPoint significantly boosts performance for semi-supervised point cloud segmentation on many state-of-the-art backbones without requiring architectural changes. Codes are available at https://github.com/jojodidli/SemiGMMPoint. •Gaussian mixture-based generative classifiers improve point cloud segmentation.•Minimizing distribution similarity can improve the performance of GMM classifiers.•The sampling bias issue under 3D semi-supervised settings can be alleviated through GMM.•Combining Generative and contrastive methods fosters semi-supervised 3D segmentation.