AI Paper Atlas

Research map / Label-efficient and robust learning / Self-supervised and contrastive

Geography-Aware Self-Supervised Learning

Kumar Ayush, Burak Uzkent, Chenlin Meng, Kumar Tanmay, Marshall Burke, David Lobell, Stefano Ermon

ICCV 2021 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Contrastive versus non-contrastive theory · 325 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
add geo-location classification pre-text task (ablation)MoCo-V2 · fMoW single-image classification (frozen features)accuracy: 60.69 -> 64.07Improved
add temporal positive pairs (ablation)MoCo-V2 · fMoW single-image classification (frozen features)accuracy: 60.69 -> 68.32Improved
add geo-location task on top of temporal positives (ablation)MoCo-V2+TP · fMoW single-image classification (frozen features)accuracy: 68.32 -> 66.33Hurt
add geo-location task and temporal positives, end-to-end finetuning (ablation)MoCo-V2 · fMoW single-image classification (finetune)accuracy: 64.34 -> 70.60Improved
pre-train MoCo-V2 on fMoW instead of ImageNet (ablation)MoCo-V2 · fMoW single-image classification (finetune)accuracy: 62.90 -> 64.34Improved
classify temporal data per area instead of single images (ablation)MoCo-V2 · fMoW temporal classification (frozen features)accuracy: 60.69 -> 68.64Improved
add temporal positives and geo-location task (ablation)MoCo-V2 · xView object detectionAP: 15.45 -> 17.74Improved
add geo-location task alone (ablation)MoCo-V2 · xView object detectionAP: 15.45 -> 15.63No clear effect
add temporal positives and geo-location task (ablation)MoCo-V2 · SpaceNet segmentationmIoU: 78.05 -> 78.51Improved
add geo-location task (ablation)MoCo-V2 · NAIP land cover classificationtop-1 accuracy: 55.18 -> 58.23Improved
add temporal positives to geo-location task (ablation)MoCo-V2+Geo · NAIP land cover classificationtop-1 accuracy: 58.23 -> 57.63Hurt
add geo-location task (ablation)MoCo-V2 · GeoImageNettop-1 accuracy: 38.51 -> 39.96Improved

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