AI Paper Atlas

Research map / 3D vision / Depth and stereo

Physics-Based Differentiable Depth Sensor Simulation

Benjamin Planche, Rajat Vikram Singh

ICCV 2021 · 3D vision › Depth and stereo · approach: Depth from light field and defocus · 10 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 3D clutter to the virtual scenes (ablation)DDS (untrained) + DR · Cropped LineMODaccuracy: 50.0 -> 89.6Improved
add 3D clutter to the virtual scenes (ablation)DDS (|Phi|=2,535) + DR · Cropped LineMODaccuracy: 54.3 -> 93.0Improved
add 3D clutter to the virtual scenes (ablation)Basic · Cropped LineMODaccuracy: 21.3 -> 46.8Improved
add online domain randomization (DR) (ablation)DDS (untrained) · Cropped LineMODaccuracy: 69.7 -> 89.6Improved
add online domain randomization (DR) (ablation)Basic · Cropped LineMODaccuracy: 46.8 -> 70.7Improved
optimize 4 simulation parameters unsupervised (ablation)DDS · Cropped LineMODaccuracy: 69.7 -> 81.2Improved
add two trainable post-processing convolutions (2,535 parameters) (ablation)DDS (|Phi|=4) · Cropped LineMODaccuracy: 81.2 -> 85.5Improved
add two trainable post-processing convolutions with DR (ablation)DDS (|Phi|=4) + DR · Cropped LineMODaccuracy: 90.5 -> 93.0Improved
add real annotated training data to DDS (ablation)DDS (|Phi|=2,535) + DR · Cropped LineMODaccuracy: 93.0 -> 97.8Improved
train on DDS data instead of real data (ablation)T trained on real data · Cropped LineMODrotation error: 35.0 -> 31.3Improved
train on DDS data instead of real data (ablation)T trained on real data · Cropped LineMODaccuracy: 95.4 -> 93.0Hurt
train DDS on scene 3 real scans (supervised) (ablation)DDS (untrained) · 2D-3D-S segmentationpixel accuracy: 62.9 -> 69.8Improved

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