Contrastive Embedding Distribution Refinement and Entropy-Aware Attention for 3D Point Cloud Classification

Abstract

Learning a powerful representation from point clouds is a fundamental and challenging problem in the field of computer vision. Different from images where RGB pixels are stored in the regular grid, for point clouds, the underlying semantic and structural information of point clouds is the spatial layout of the points. Moreover, the properties of challenging incontext and background noise pose more challenges to point cloud analysis. One assumption is that the poor performance of the classification model can be attributed to the indistinguishable embedding feature that impedes the search for the optimal classifier. This work offers a new strategy for learning powerful representations via a contrastive learning approach that can be embedded into any point cloud classification network. First, we propose a supervised contrastive classification (SCC) method to implement embedding feature distribution refinement by improving the intra-class compactness and inter-class separability. Second, to solve the confusion problem caused by small inter-class variations between some similar-looking categories, we propose a confusion-prone class mining (CPCM) strategy to alleviate the confusion effect. Finally, considering that outliers of the sample clusters in the embedding space may cause performance degradation, we design an entropy-aware attention (EAA) module with information entropy theory to identify the outlier cases and the unstable samples by measuring the uncertainty of predicted probability. We assume that the outlier and unstable samples need to be ignored and enhanced, respectively, to enable the model to learn more robust decision boundaries. The results of extensive experiments demonstrate that our method outperforms the state-of-the-art approaches by achieving 82.9% accuracy on the real-world ScanObjectNN dataset and substantial performance gains up to 2.9% in DCGNN, 3.1% in PointNet++, and 2.4% in GBNet. The code will be available at https://github.com/YangFengSEU/CEDR.

MFTR-Net: A Multi-Level Features Network with Targeted Regularization for Large-Scale Point Cloud Classification

“…After feature extraction in a point cloud classification task, it can accelerate feature propagation. Yang et al [20] proposed a supervised contrastive point cloud classification method to implement embedding feature distribution refinement by improving intra-class compactness and inter-class separability, which solves the confusion problem caused by slight inter-class variations and the confusion problem caused by small inter-class compactness and inter-class separability. Zhang et al [12] proposed a method that automatically learns a data augmentation strategy using bilevel optimization, minimizing a base model's loss on a validation set when the augmented input is used for training the model.…”

Uncovering the Origins of Instability in Dynamical Systems: How Can the Attention Mechanism Help?

“…Some studies have tried to check the stability properties of graph neural networks to see how changes in the underlying topology can affect the output of the network [7]. In terms of model optimization, it has been discussed that unstable nodes in sparse regions of the network require to be pulled apart to improve the classification decision [8].…”