Evolutionary Neural Architecture Search for 3D Point Cloud Analysis

TL;DR

This paper introduces SHSADE-PIDS, an evolutionary NAS framework that efficiently designs neural architectures for 3D point cloud analysis, achieving state-of-the-art accuracy with minimal parameters and computational cost.

Contribution

It presents a novel evolutionary NAS method that encodes architectures into continuous spaces and applies it to 3D point cloud tasks, outperforming prior techniques.

Findings

Achieved 64.51% mean IoU on SemanticKITTI with 0.55M parameters.

Surpassed state-of-the-art accuracy on ModelNet40 with 93.4%.

Reduced computational overhead by over 22-26 times.

Abstract

Neural architecture search (NAS) automates neural network design by using optimization algorithms to navigate architecture spaces, reducing the burden of manual architecture design. While NAS has achieved success, applying it to emerging domains, such as analyzing unstructured 3D point clouds, remains underexplored due to the data lying in non-Euclidean spaces, unlike images. This paper presents Success-History-based Self-adaptive Differential Evolution with a Joint Point Interaction Dimension Search (SHSADE-PIDS), an evolutionary NAS framework that encodes discrete deep neural network architectures to continuous spaces and performs searches in the continuous spaces for efficient point cloud neural architectures. Comprehensive experiments on challenging 3D segmentation and classification benchmarks demonstrate SHSADE-PIDS's capabilities. It discovered highly efficient architectures with higher accuracy, significantly advancing prior NAS techniques. For segmentation on SemanticKITTI, SHSADE-PIDS attained 64.51% mean IoU using only 0.55M parameters and 4.5GMACs, reducing overhead by over 22-26X versus other top methods. For ModelNet40 classification, it achieved 93.4% accuracy with just 1.31M parameters, surpassing larger models. SHSADE-PIDS provided valuable insights into bridging evolutionary algorithms with neural architecture optimization, particularly for emerging frontiers like point cloud learning.