AI Paper Atlas

Research map / 3D vision / Face and shape

DebSDF: Delving Into the Details and Bias of Neural Indoor Scene Reconstruction

Yuting Xiao, Jingwei Xu, Zehao Yu, Shenghua Gao

TPAMI 2024 · 3D vision › Face and shape · approach: Neural implicit surface reconstruction · 17 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 uncertainty-guided prior filtering (ablation)MonoSDF · ScanNetF-score: 73.30 -> 76.82Improved
add uncertainty-guided prior filtering (ablation)MonoSDF · ICL-NUIMF-score: 67.39 -> 73.97Improved
add uncertainty-guided ray sampling (ablation)DebSDF with prior filtering · ScanNetF-score: 76.82 -> 77.12Improved
add uncertainty-guided ray sampling (ablation)DebSDF with prior filtering · ICL-NUIMF-score: 73.97 -> 76.14Improved
add uncertainty-guided smooth regularization (ablation)DebSDF with filtering and sampling · ScanNetF-score: 77.12 -> 77.30Improved
add uncertainty-guided smooth regularization (ablation)DebSDF with filtering and sampling · ICL-NUIMF-score: 76.14 -> 76.75Improved
add bias-aware SDF-to-density transformation (ablation)DebSDF with three uncertainty modules · ScanNetF-score: 77.30 -> 78.54Improved
add bias-aware SDF-to-density transformation (ablation)DebSDF with three uncertainty modules · ICL-NUIMF-score: 76.75 -> 77.59Improved
predict uncertainty with view-dependent color network instead of geometry network (ablation)DebSDF · ScanNetF-score: 73.2 -> 78.5Improved
fuse depth and normal uncertainty with lambda 0.9 instead of lambda 0 (depth only) (ablation)DebSDF · ICL-NUIMF-score: 75.67 -> 77.39Improved
use lambda 0.9 instead of 1 (normal only) (ablation)DebSDF · ICL-NUIMF-score: 77.16 -> 77.39Improved
replace SDF-to-density transformation with bias-aware one (ablation)TUVR / NeuS · DTU (no prior)Chamfer: 0.79 -> 0.75Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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