AI Paper Atlas

Research map / 3D vision / Depth and stereo

Towards Robust Monocular Depth Estimation: Mixing Datasets for Zero-Shot Cross-Dataset Transfer

Rene Ranftl, Katrin Lasinger, David Hafner, Konrad Schindler B, Vladlen Koltun

TPAMI 2020 · 3D vision › Depth and stereo · approach: Self-supervised monocular depth · 2,295 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
Use Pareto-optimal multi-objective mixing instead of naive mixing (all five datasets) (ablation)MIX 5 · Mean over 6 zero-shot test setsmean relative performance [%]: 19.48 -> 22.40Improved
Add MV to MIX 1 with naive mixing (MIX 1 to MIX 2) (ablation)naive mixing · Mean over 6 zero-shot test setsmean relative performance [%]: 18.26 -> 17.34Hurt
Add MV to MIX 1 with Pareto mixing (MIX 1 to MIX 2) (ablation)Pareto mixing · Mean over 6 zero-shot test setsmean relative performance [%]: 16.56 -> 19.75Improved
Add MD to MIX 2 with Pareto mixing (MIX 3) (ablation)Pareto mixing · Mean over 6 zero-shot test setsmean relative performance [%]: 19.75 -> 20.79Improved
Add WS to MIX 3 (MIX 5) with Pareto mixing (ablation)Pareto mixing · Mean over 6 zero-shot test setsmean relative performance [%]: 20.79 -> 22.40Improved
Add 3D Movies (MV) to MIX 4 (giving MIX 5), Pareto mixing (ablation)Pareto mixing · Mean over 6 zero-shot test setsmean relative performance [%]: 21.6 -> 22.40Improved
Fine-tune on single dataset DL instead of RW only (ablation)RW baseline · Mean over 6 zero-shot test setsmean relative performance [%]: -10.20Hurt
Fine-tune on single dataset WS instead of RW only (ablation)RW baseline · Mean over 6 zero-shot test setsmean relative performance [%]: -21.37Hurt
Swap ResNet-50 encoder for ResNeXt-101 pretrained on weakly-supervised data (ablation)MiDaS depth network · Validation sets (RW, MD, MV)relative performance: up to 15 % relative improvementImproved
Initialize encoder with ImageNet pretraining vs random init (ablation)ResNet-50 encoder · Validation setsrelative performance: 35\% worseImproved
Use trimmed MAE scale-shift-invariant loss L_ssitrim + L_reg (ablation)depth network trained on RW · Validation setsrelative performanceImproved
Use full (ResNeXt-101-WSL) model instead of small ResNet-50 modelOurs · Mean over 6 zero-shot test setsmean relative performance [%]: -16.0Improved

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