AI Paper Atlas

Research map / 3D vision / Depth and stereo

Bridging the Domain Gap for Ground-to-Aerial Image Matching

Krishna Regmi, Mubarak Shah

ICCV 2019 · 3D vision › Depth and stereo · approach: Scene coordinate regression for localization · 204 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 joint feature learning with synthesized aerial imagestwo-stream (Ig, Ia) · CVUSAtop-1% recall: 82.94 -> 92.09Improved
add feature fusion of synthesized-aerial and ground featuresJoint Feature Learning · CVUSAtop-1 recall: 29.75 -> 48.75Improved
add feature fusionJoint Feature Learning · CVUSAtop-1% recall: 92.09 -> 95.98Improved
use synthesized aerial image as query instead of real ground imagetwo-stream · CVUSAtop-1% recall: 82.94 -> 72.58Hurt
add multi-scale features (conv_6, conv_7 with conv_8) (ablation)Feature Fusion · CVUSAtop-1 recall: 8.01 -> 48.75Improved
do not use global average pooling (ablation)Feature Fusion · CVUSAtop-1 recall: 16.13 -> 48.75Improved
do not share weights between ground and aerial encoders (ablation)Feature Fusion · CVUSAtop-1 recall: 29.94 -> 48.75Improved
do not share weights between encoders (ablation)Feature Fusion · CVUSAtop-1% recall: 93.42 -> 95.98Improved
feature fusion vs CVM-Net-IFeature Fusion · CVUSAtop-1% recall: 91.4 -> 95.98Improved
add feature fusion on urban OP datasetJoint Feature Learning · Orlando-Pittsburgh (OP)top-1 recall: 38.36 -> 45.57Improved
add joint feature learning on urban OP datasettwo-stream (Ig, Ia) · Orlando-Pittsburgh (OP)top-1 recall: 30.61 -> 38.36Improved
add feature fusion for reverse aerial-to-ground matchingJoint Feature Learning · CVUSA aerial-to-groundtop-1 recall: 27.39 -> 44.99Improved

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