AI Paper Atlas

Research map / 3D vision / Depth and stereo

Digging Into Uncertainty in Self-Supervised Multi-View Stereo

Hongbin Xu, Zhipeng Zhou, Yali Wang, Wenxiong Kang, Baigui Sun, Hao Li, Yu Qiao

ICCV 2021 · 3D vision › Depth and stereo · approach: Foundation models for monocular geometry · 72 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 flow-depth consistency loss L_fc (ablation)U-MVS (MVSNet) · DTU evaluationoverall error: 0.5436 -> 0.4820Improved
add uncertainty-aware self-training loss L_uc (ablation)U-MVS (MVSNet) · DTU evaluationoverall error: 0.4820 -> 0.4501Improved
add flow-depth consistency loss L_fc (ablation)U-MVS (CascadeMVSNet) · DTU evaluationoverall error: 0.4041 -> 0.3789Improved
add uncertainty-aware self-training loss L_uc (ablation)U-MVS (CascadeMVSNet) · DTU evaluationoverall error: 0.3789 -> 0.3537Improved
add flow-depth consistency loss L_fc (ablation)U-MVS (CascadeMVSNet) · DTU evaluationcompleteness error: 0.3641 -> 0.3833Hurt
add flow-depth consistency loss L_fc (ablation)U-MVS (CascadeMVSNet) · DTU evaluationaccuracy error: 0.4442 -> 0.3745Improved
add L_fc and L_uc to photometric loss (ablation)U-MVS (MVSNet) · DTU evaluationaccuracy error: 0.5527 -> 0.4695Improved
self-supervised U-MVS instead of supervised training with same CascadeMVSNet backboneCascadeMVSNet · Tanks and Temples intermediatemean F-score: 56.42 -> 57.15Improved
self-supervised U-MVS instead of supervised training with same CascadeMVSNet backboneCascadeMVSNet · Tanks and Temples advancedmean F-score: 31.12 -> 30.97No clear effect
self-supervised U-MVS instead of supervised training with same CascadeMVSNet backboneCascadeMVSNet · DTU evaluationoverall error: 0.355 -> 0.3537No clear effect
self-supervised U-MVS instead of supervised training with same CascadeMVSNet backboneCascadeMVSNet · DTU evaluationaccuracy error: 0.325 -> 0.354Hurt
self-supervised U-MVS instead of supervised training with same CascadeMVSNet backboneCascadeMVSNet · DTU evaluationcompleteness error: 0.385 -> 0.3535Improved

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