SAGE3D: Soft-guided attention and graph excitation for 3D point cloud corner detection

TL;DR

SAGE3D is a Transformer-based model that improves 3D corner detection in LiDAR point clouds using hierarchical architecture, soft-guided attention, and excitatory graph neural networks for enhanced precision and recall.

Contribution

The paper introduces Soft-Guided Attention and Excitatory Graph Neural Networks within a hierarchical framework for more accurate 3D corner detection.

Findings

Improved corner detection precision through guided attention.

Enhanced recall via excitatory graph neural networks.

Hierarchical architecture enables effective multi-scale feature extraction.

Abstract

We present SAGE3D, a hybrid Transformer-based model for corner detection in airborne LiDAR point clouds. We propose a multi-stage solution built on a hierarchical encoder-decoder architecture that progressively downsamples point clouds through Set Abstraction layers and recovers per-point predictions via Feature Propagation. We introduce two innovations: Soft-Guided Attention, which injects ground-truth corner labels as a log-prior into attention logits during training to improve precision; then an Excitatory Graph Neural Network positioned at strategic resolutions in the hierarchy, employing positive-only message passing where high-confidence corners reinforce predictions through learned boosting, optimizing for recall. The hierarchical design enables multi-scale feature extraction while our guided attention and excitatory modules ensure corner signals are amplified rather than diluted across scales.