AI Paper Atlas

Research map / 3D vision / Depth and stereo

GeoNet++: Iterative Geometric Neural Network with Edge-Aware Refinement for Joint Depth and Surface Normal Estimation

Xiaojuan Qi, Zhengzhe Liu, Renjie Liao, Philip H. S. Torr, Raquel Urtasun, Jiaya Jia

TPAMI 2020 · 3D vision › Depth and stereo · approach: Dense depth prediction and completion · 78 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 GeoNet depth-to-normal and normal-to-depth modules (ablation)VGG-16 baseline · NYUD-V2 normalmean angle error (deg): 19.4 -> 19.0Improved
add ensemble module (ablation)GeoNet · NYUD-V2 normalmean angle error (deg): 19.0 -> 18.9Improved
add edge-aware refinement (ablation)GeoNet · NYUD-V2 normalmean angle error (deg): 19.0 -> 18.6Improved
add iterative inference (full model) (ablation)GeoNet++ · NYUD-V2 normalmean angle error (deg): 18.6 -> 18.5Improved
add GeoNet modules (ablation)VGG-16 baseline · NYUD-V2 depthrmse: 0.626 -> 0.608Improved
add full GeoNet++ modules (ablation)VGG-16 baseline · NYUD-V2 depthrmse: 0.626 -> 0.600Improved
add GeoNet++ modules (ablation)VGG-16 baseline · NYUD-V2 3DGMmean angle error (deg): 42.39 -> 33.24Improved
use explicit geometric modules instead of geometric lossVGG-16 baseline · NYUD-V2 depthrmse: 0.615 -> 0.600Improved
change Canny thresholds to extremely low values (ablation)GeoNet++ · NYUD-V2 normalmean angle error (deg): 18.6 -> 18.8No clear effect
change Canny thresholds to extremely high values (ablation)GeoNet++ · NYUD-V2 depthrmse: 0.605 -> 0.605No clear effect
add GeoNet++ refinement on top of DORNDORN · KITTI depthrmse: 4.06 -> 4.10Hurt
add GeoNet++ refinement on top of DORNDORN · NYUD-V2 depthrmse: 0.552 -> 0.527Improved

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