PointViG: A Lightweight GNN-based Model for Efficient Point Cloud Analysis

TL;DR

This study introduces PointViG, an efficient framework for point cloud analysis that incorporates a lightweight graph convolutional module to efficiently aggregate local features and mitigate over-smoothing and proposes an adaptive dilated graph convolution technique that searches for sparse neighboring nodes within a dilated neighborhood based on semantic correlation.

Abstract

In the domain of point cloud analysis, despite the significant capabilities of Graph Neural Networks (GNNs) in managing complex 3D datasets, existing approaches encounter challenges like high computational costs and scalability issues with extensive scenarios. These limitations restrict the practical deployment of GNNs, notably in resource-constrained environments. To address these issues, this study introduce <b>Point<\b> <b>Vi<\b>sion <b>G<\b>NN (PointViG), an efficient framework for point cloud analysis. PointViG incorporates a lightweight graph convolutional module to efficiently aggregate local features and mitigate over-smoothing. For large-scale point cloud scenes, we propose an adaptive dilated graph convolution technique that searches for sparse neighboring nodes within a dilated neighborhood based on semantic correlation, thereby expanding the receptive field and ensuring computational efficiency. Experiments demonstrate that PointViG achieves performance comparable to state-of-the-art models while balancing performance and complexity. On the ModelNet40 classification task, PointViG achieved 94.3% accuracy with 1.5M parameters. For the S3DIS segmentation task, it achieved an mIoU of 71.7% with 5.3M parameters. These results underscore the potential and efficiency of PointViG in point cloud analysis.

Figures (10)

• Figure 1: Comparison of classification accuracy among representative models, where bubble areas correspond to the number of floating-point operations (FLOPs). The specific FLOPs values (in billions) are provided in textual notation. PointViG surpasses other models, demonstrating superior performance with fewer parameters and FLOPs. This highlights PointViG's optimal trade-off between performance and complexity.
• Figure 2: PointViG Module acts as the basic module in PointViG framework.
• Figure 3: The PointViG architecture designed for classification.
• Figure 4: The PointViG architecture designed for semantic segmentation.
• Figure 5: Illustration of the feature diversity output by each layer within the three PointViG Modules utilized in classification models.