Human Vision Based 3D Point Cloud Semantic Segmentation of Large-Scale Outdoor Scene

TL;DR

This paper introduces EyeNet, a novel human vision-inspired network for large-scale outdoor 3D point cloud semantic segmentation, improving coverage handling and achieving state-of-the-art results.

Contribution

The paper presents a new network architecture that incorporates multi-contour inputs and parallel streams with connection blocks, inspired by human peripheral vision.

Findings

Achieves state-of-the-art performance on large-scale outdoor datasets.

Effectively handles dense point clouds with improved coverage.

Demonstrates through ablation studies the effectiveness of the proposed components.

Abstract

This paper proposes EyeNet, a novel semantic segmentation network for point clouds that addresses the critical yet often overlooked parameter of coverage area size. Inspired by human peripheral vision, EyeNet overcomes the limitations of conventional networks by introducing a simple but efficient multi-contour input and a parallel processing network with connection blocks between parallel streams. The proposed approach effectively addresses the challenges of dense point clouds, as demonstrated by our ablation studies and state-of-the-art performance on Large-Scale Outdoor datasets.