Semantic Segmentation for Real Point Cloud Scenes via Bilateral Augmentation and Adaptive Fusion

TLDR

Fine‑grained analysis of basic point‑cloud data is warranted because real‑world scenes capture complex surroundings yet raw 3D data pose significant perception challenges. The study aims to perform semantic segmentation on large‑scale, real‑world point‑cloud data. The authors augment local context with a bilateral structure that fuses geometric and semantic cues, then apply an adaptive, multi‑resolution fusion at the point level, and validate the modules with ablation studies and visualizations. Comparisons with state‑of‑the‑art networks on three benchmarks show that the proposed network is effective.

Abstract

Given the prominence of current 3D sensors, a fine-grained analysis on the basic point cloud data is worthy of further investigation. Particularly, real point cloud scenes can intuitively capture complex surroundings in the real world, but due to 3D data’s raw nature, it is very challenging for machine perception. In this work, we concentrate on the essential visual task, semantic segmentation, for large-scale point cloud data collected in reality. On the one hand, to reduce the ambiguity in nearby points, we augment their local context by fully utilizing both geometric and semantic features in a bilateral structure. On the other hand, we comprehensively interpret the distinctness of the points from multiple resolutions and represent the feature map following an adaptive fusion method at point-level for accurate semantic segmentation. Further, we provide specific ablation studies and intuitive visualizations to validate our key modules. By comparing with state-of-the-art networks on three different benchmarks, we demonstrate the effectiveness of our network.