AI Paper Atlas

Research map / Large language models / AI and society

A Benchmark Dataset for Spatially Aligned Road Damage Assessment in Small Uncrewed Aerial Systems Disaster Imagery

Thomas Manzini, Priyankari Perali, Raisa Karnik, Robin R. Murphy

AAAI 2026 · Large language models › AI and society · approach: Benchmarking and reproducibility practice

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use spatially aligned (adjusted) road lines instead of unadjusted (ablation)8 baseline models · Simple taskIoU_km: average degradation 0.014 IoU_km when unadjustedImproved
use adjusted road lines on Full task (ablation)8 baseline models · Full taskIoU_km: average degradation 0.0003 IoU_kmNo clear effect
use adjusted road lines on UNet with Attention (ablation)UNet with Attention · Simple taskIoU_km: 0.312 -> 0.331Improved
add attention to UNet (ablation)UNet · Simple taskmacro IoU_km: 0.307 -> 0.331Improved
use pretrained ViT encoder with Segmenter (ablation)ViT-L + Segmenter · Simple taskmacro IoU_km: 0.039 -> 0.015Hurt
use pretrained ViT encoder with UperNet (ablation)ViT-L + UperNet · Simple taskmacro IoU_km: 0.211 -> 0.209No clear effect
increase number of prediction classes from Simple to FullAll baseline models · CRASAR-U-DROIDsmacro IoU_km: 0.331 -> 0.091 (UNet with Attention)Hurt
deploy Attention UNet Simple model in Hurricanes Debby and HeleneUNet with Attention · operational deploymentqualitative practitioner feedbackImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy