AI Paper Atlas

Research map / Detection and segmentation / Open-set and few-shot detection

Caltech Aerial RGB-Thermal Dataset in the Wild

Connor Lee, Matthew Anderson, Nikhil Ranganathan, Xingxing Zuo, Kevin T Do, Georgia Gkioxari, Soon-Jo Chung

ECCV 2024 · Detection and segmentation › Open-set and few-shot detection · approach: Remote sensing segmentation and change · 24 citations (OpenAlex)

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 larger model (ConvNext-B) instead of fastest FastSCNNthermal semantic segmentation · general splitmIoU: +0.027 mIoU at 130x FLOPsNo clear effect
use frozen large RGB foundation models (DINOv2, ConvNext-B CLIP) instead of ImageNet-pretrained lightweight netsthermal segmentation · general splitmIoUNo clear effect
use thermal-specific FTNet edge loss instead of cross entropy baselinesthermal segmentation · general splitmIoUHurt
train on one region/terrain and test on others (ablation)EfficientViT-B0 · geographic splitavg mIoU: 0.314 (train CA) vs 0.586 (train All)Hurt
add ground-captured images to aerial training data (ablation)EfficientViT-B0 · terrain splitavg mIoU: 0.518 -> 0.532Improved
train on daytime only and test at sunrise or night (ablation)EfficientViT-B0 · temporal splitmIoU: 0.777 (day) -> 0.242 (sunrise), 0.193 (night)Hurt
add RGB to thermal input in RGB-T fusion modelssemantic segmentation · RGB-T splitmIoU: 0.661 (thermal only) -> 0.740 (CMNeXt)Improved
use paired GANs instead of unpaired GANs or diffusion for RGB-to-thermal translationimage translation · RGB to thermalPSNR/SSIM/mIoU: thermal mIoU 0.189 (UNIT) -> 0.388 (VQGAN)Improved
use thermal imagery instead of RGB for SAM and vision-language zero-shot segmentationfoundation models · zero-shotmIoU / AP: ODISE mIoU 0.504 (color) -> 0.232 (thermal); SAM AP 0.538 -> 0.018Hurt
pretrain with MoCoV2 on urban thermal data instead of ImageNet weights (ablation)thermal segmentation networks · Caltech Aerial test setmIoU: EfficientViT 0.725 -> 0.714No clear effect
use ImageNet pretraining instead of training from scratch (ablation)thermal segmentation networks · Caltech Aerial test setmIoU: EfficientViT 0.687 -> 0.725; FastSCNN 0.640 -> 0.690Improved
use thermal instead of RGB input for VINS-FusionVINS-Fusion · aerial sequencesATE: North Field N1 2.883 (RGB) vs 9.454 (thermal); Castaic Lake C2 --- vs 1.513Mixed

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