AI Paper Atlas

Research map / 3D vision / Depth and stereo

Camera Height Doesn't Change: Unsupervised Training for Metric Monocular Road-Scene Depth Estimation

Genki Kinoshita, Ko Nishino

ECCV 2024 · 3D vision › Depth and stereo · approach: Self-supervised monocular depth · 9 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 FUMET training to Monodepth2Monodepth2 · KITTI Eigen (median scaling)AbsRel: 0.115 -> 0.106Improved
optimize camera height across frames and epochs rather than per-frame size prior (ablation)FUMET · KITTIAbsRel: 0.115 -> 0.108Improved
add outlier filtering in Silhouette Projector (ablation)FUMET · KITTIAbsRel: 0.116 -> 0.108Improved
balance loss weights when only one of aux or cam loss is used (ablation)FUMET · KITTIAbsRel: 0.125 -> 0.149 (aux only); 0.116 -> 0.211 (cam only)Hurt
balance loss weights when both aux and cam loss are used (ablation)FUMET · KITTIAbsRel: 0.122 -> 0.108Improved
train FUMET on mixed datasets plus KITTI instead of KITTI onlyFUMET · KITTI Eigen (832x512)AbsRel: 0.103 -> 0.082Improved
fix camera height computed offline from pre-trained model instead of online optimization (ablation)FUMET · KITTIAbsRel: 0.108 -> 0.106Improved
add YouTube in-the-wild videos to mixed plus KITTI trainingFUMET · KITTI Eigen (832x512)AbsRel: 0.082 -> 0.084No clear effect
use recovered COLMAP intrinsics instead of ground-truth intrinsicsFUMET · KITTIAbsRel: 0.108 -> 0.110No clear effect
use FUMET without depth supervision versus supervised ZoeDepth on mixed dataFUMET · KITTI Eigen (832x512)AbsRel: 0.120 -> 0.113Improved
add occlusion and truncation augmentations when training LSP (ablation)LSP · KITTI car heightabsolute relative error: 0.081 -> 0.062Improved
predict car width and length as well as height in LSP (ablation)LSP · KITTI car height (All)absolute relative error: 0.087 -> 0.062Improved

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