AI Paper Atlas

Research map / 3D vision / Depth and stereo

Learning Monocular Depth by Distilling Cross-domain Stereo Networks

Xiaoyang Guo, Hongsheng Li, Shuai Yi, Jimmy Ren, Xiaogang Wang

ECCV 2018 · 3D vision › Depth and stereo · approach: Self-supervised monocular depth · 210 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
distill from a synthetic-pretrained stereo network instead of training monocular net directly on synthetic datamonocular depth network · KITTI Eigen (80m)Abs Rel: 0.662 -> 0.109Improved
pretrain monocular network on Scene Flow before KITTI supervisionDirect Sup. monocular · KITTI Eigen (all of K)Abs Rel: 0.105 -> 0.106No clear effect
pretrain monocular network on Scene Flow with 100 KITTI labelsDirect Sup. monocular · KITTI Eigen (100 of K)Abs Rel: 0.187 -> 0.194Hurt
add unsupervised fine-tuning of the stereo proxy (ours)StereoNoFt->Mono · KITTI Eigen (80m)Abs Rel: 0.109 -> 0.105Improved
add ImageNet pretraining of monocular encoderStereoUnsupFt->Mono · KITTI Eigen (80m)Abs Rel: 0.105 -> 0.099Improved
add Cityscapes pretraining stage for monocular networkStereoUnsupFt->Mono pt · KITTI Eigen (80m)Abs Rel: 0.099 -> 0.095Improved
use our occlusion-aware unsupervised loss instead of Godard et al. loss for stereo fine-tuningStereoUnsupFt · KITTI Eigen stereoAbs Rel: 0.076 -> 0.061Improved
use our occlusion-aware unsupervised loss instead of Godard et al. lossStereoUnsupFt->Mono · KITTI Eigen distilled monoAbs Rel: 0.116 -> 0.105Improved
use Godard et al. unsupervised loss to finetune stereo net instead of no fine-tuningStereoNoFt · KITTI Eigen stereoAbs Rel: 0.072 -> 0.076Hurt
add supervised fine-tuning with 100 images to the stereo netStereoNoFt · KITTI Eigen stereoRMS: 3.836 -> 3.325Improved
add unsupervised fine-tuning to the stereo netStereoNoFt · Scene Flow testMAE (px): 3.41 -> 4.23Hurt
add supervised 100-image fine-tuning to stereo before distillationStereoNoFt->Mono · KITTI Eigen distilled monoRMS: 4.656 -> 4.449Improved

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