Aaron E. Maxwell

Comparing UNet and HRNet for Geomorphic Feature Extraction from Digital Terrain Model-Derived Land Surface Parameters

IntELiMon: an automated framework for standardized terrestrial lidar processing in multi-disciplinary ecosystem monitoring

Terrestrial laser scanning (TLS) enables rapid, objective, and repeatable measurement of 3D forest and shrubland structure, but operational use is often limited by complex, multi-stage processing workflows. IntELiMon (Interagency Ecosystem lidar Monitoring) is a flexible, automated processing system that streamlines...

Comparing feature maps generated using UNet-like CNN, 2 Transformer, Mamba, and hybrid architectures for general

This study compares feature maps produced by semantic segmentation architectures using varying combinations of convolutional neural network (CNN), Transformer, and Mamba selective 26 state space (selective SSM) components with a goal of exploring the following question: does correlation or similarity between the genera...

SegFormer and SegFormer-UNet for anthropogenic geomorphic feature extraction from land surface parameters

Three architectures for pixel-level extraction of features from light detection and ranging (lidar)-derived land surface parameters (LSPs) are compared: SegUNet and the UNet typically achieve the best precision and recall balance and the smoothest optimization, and UNet-style architectures with CNN-based decoders gener...