AI Paper Atlas

Research map / 3D vision / Depth and stereo

MVSNet: Depth Inference for Unstructured Multi-view Stereo

Yao Yao, Zixin Luo, Shiwei Li, Tian Fang, Long Quan

ECCV 2018 · 3D vision › Depth and stereo · approach: Cost-volume multi-view stereo · 1,634 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 depth map refinement network (ablation)MVSNet · DTU evaluation setf-score (<1mm): 75.58 -> 75.69Improved
add depth map refinement network (ablation)MVSNet · DTU evaluation setf-score (<2mm): 79.98 -> 80.25Improved
add depth map refinement network (ablation)MVSNet · DTU validationvalidation lossNo clear effect
use learned 2D feature extraction network instead of single 7x7 stride-4 conv layer (ablation)MVSNet · DTU validationvalidation lossImproved
use variance-based cost metric instead of mean (ablation)MVSNet · DTU validationvalidation lossImproved
increase test input views from 3 to 5 (trained with 3) (ablation)MVSNet · DTU validationvalidation lossImproved
add more input views (N=2 vs 3) (ablation)MVSNet · DTU validationvalidation lossImproved
use MVSNet instead of Gipuma on completenessMVSNet · DTU evaluation setcompleteness distance (mm): 0.873 -> 0.527Improved
use MVSNet instead of Gipuma on accuracyMVSNet · DTU evaluation setaccuracy distance (mm): 0.283 -> 0.396Hurt
use MVSNet instead of Gipuma overallMVSNet · DTU evaluation setoverall distance (mm): 0.578 -> 0.462Improved
apply DTU-trained MVSNet to outdoor scenes without fine-tuningMVSNet · Tanks and Temples intermediatemean f-score: 42.14 -> 43.48Improved

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