AI Paper Atlas

Research map / Vision-language and multimodal / CLIP and zero-shot

WildSAT: Learning Satellite Image Representations from Wildlife Observations

Rangel Daroya, Elijah Cole, Oisin Mac Aodha, Grant Van Horn, Subhransu Maji

ICCV 2025 · Vision-language and multimodal › CLIP and zero-shot · approach: Adapting VLMs for zero-shot classification · 16 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
fine-tune with WildSAT on species observationsbase models (avg across 15 architectures, no random) · UCM linear probingaccuracy: 81.8 -> 87.9Improved
fine-tune with WildSAT on species observationsbase models (avg incl. random) · BEN20k linear probingmicro F1: 38.5 -> 53.1Improved
fine-tune with WildSATrandomly initialized ViT-B/16 · BEN20k linear probingmicro F1: 0.0 -> 51.2Improved
fine-tune with WildSATCLIP ViT-B/16 · 7 satellite datasets (avg)accuracy: 71.6% -> 76.6%Improved
fine-tune CLIP with cross-modal methods (TaxaBind, GRAFT) instead of leaving unchangedCLIP ViT-B/16 · 7 satellite datasets (avg)accuracy: 71.6% -> 59.8%Hurt
add WildSAT fine-tuningSatlasNet, SatMAE, TaxaBind · SatBird USA Summertop-k accuracy: 47.66 -> 49.53Improved
add WildSAT fine-tuningSatlasNet, SatMAE, TaxaBind · SatBird Kenyatop-k accuracy: 23.80 -> 24.12Improved
add WildSAT fine-tuningResNet50 encoders · Cashew1k segmentationIoU: 59.9% -> 72.2%Improved
add WildSAT fine-tuningResNet50 encoders · SAcrop3k segmentationIoU: 21.4% -> 22.7%Improved
add location (loc) loss (ablation)WildSAT (random ResNet50 + ImageNet ViT-B/16) · UCM/AID/RESISC45/So2Sat20k avgaccuracy: 48.6% -> 57.5%Improved
add environmental covariates to location encoder (ablation)WildSAT with loc · UCM/AID/RESISC45/So2Sat20k avgaccuracy: 57.5% -> 62.0%Improved
add text loss (ablation)WildSAT with loc+env · UCM/AID/RESISC45/So2Sat20k avgaccuracy: 62.0% -> 66.1%Improved

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