NL-LinkNet: Toward Lighter But More Accurate Road Extraction With Nonlocal Operations
TL
TL;DR
This letter proposes an efficient nonlocal LinkNet with nonlocal blocks (NLBs) that can grasp relations between global features that enables each spatial feature point to refer to all other contextual information and results in more accurate road segmentation.
Abstract
Road extraction from very high resolution (VHR) satellite images is one of the most important topics in the field of remote sensing. In this letter, we propose an efficient nonlocal LinkNet with nonlocal blocks (NLBs) that can grasp relations between global features. This enables each spatial feature point to refer to all other contextual information and results in more accurate road segmentation. In detail, our single model without any postprocessing like conditional random field (CRF) refinement performed better than any other published state-of-the-art ensemble model in the official DeepGlobe Challenge. Moreover, our nonlocal LinkNet (NL-LinkNet) beat the D-LinkNet, the winner of the DeepGlobe challenge (Demir et al., 2018), with 43% less parameters, less giga floating-point operations per seconds (GFLOPs), and shorter training convergence time. We also present empirical analyses on the proper usages of NLBs for the baseline model.
References35 items
1
Residual Non-local Attention Networks for Image Restoration
2
Attentional Road Safety Networks
3
Parameter-Free Spatial Attention Network for Person Re-Identification
4
Visual Spatial Attention Network for Relationship Detection
5