GPr-Net: Geometric Prototypical Network for Point Cloud Few-Shot Learning

TL;DR

GPr-Net introduces a lightweight geometric prototypical network for point cloud few-shot learning, effectively capturing intrinsic topology and geometry, leading to superior performance and computational efficiency on ModelNet40.

Contribution

The paper proposes GPr-Net, a novel geometric prototypical network utilizing intrinsic geometry interpreters and hyperbolic space, advancing efficient and accurate point cloud few-shot learning.

Findings

Outperforms state-of-the-art methods on ModelNet40

Achieves 170x better computational efficiency

Effectively handles intra and inter-class variance

Abstract

In the realm of 3D-computer vision applications, point cloud few-shot learning plays a critical role. However, it poses an arduous challenge due to the sparsity, irregularity, and unordered nature of the data. Current methods rely on complex local geometric extraction techniques such as convolution, graph, and attention mechanisms, along with extensive data-driven pre-training tasks. These approaches contradict the fundamental goal of few-shot learning, which is to facilitate efficient learning. To address this issue, we propose GPr-Net (Geometric Prototypical Network), a lightweight and computationally efficient geometric prototypical network that captures the intrinsic topology of point clouds and achieves superior performance. Our proposed method, IGI++ (Intrinsic Geometry Interpreter++) employs vector-based hand-crafted intrinsic geometry interpreters and Laplace vectors to extract and evaluate point cloud morphology, resulting in improved representations for FSL (Few-Shot Learning). Additionally, Laplace vectors enable the extraction of valuable features from point clouds with fewer points. To tackle the distribution drift challenge in few-shot metric learning, we leverage hyperbolic space and demonstrate that our approach handles intra and inter-class variance better than existing point cloud few-shot learning methods. Experimental results on the ModelNet40 dataset show that GPr-Net outperforms state-of-the-art methods in few-shot learning on point clouds, achieving utmost computational efficiency that is $170\times$ better than all existing works. The code is publicly available at https://github.com/TejasAnvekar/GPr-Net.