AI Paper Atlas

Research map / 3D vision / 3D generation

3D Semantic Scene Completion: A Survey

Luis Roldão, Raoul de Charette, Anne Verroust-Blondet

IJCV 2022 · 3D vision › 3D generation · approach: Semantic scene completion · 118 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 selective middle fusion instead of simple sum/concat/max fusion (ablation)SSC networks with multi-modal input · SSC (literature)performance: at least a 20% performance boostImproved
add late fusion of 3D guidance and semantic completion branchesSSC network · SSC (literature)performance: 5 to 10%Improved
add late fusion of color and geometry branchesSSC network · SSC (literature)performance: less than 3%Improved
add semantic-wise late fusion of 2D top-view and 3D SSC (ablation)S3CNet · SemanticKITTI (outdoor)overall performance: 20% boostImproved
add CRF refinement to SSCNetSSCNet · SSC (literature)performance: little over 4% gainImproved
add Position Aware loss (ablation)PALNet · SSC (literature)performance: 1--3%Improved
add adversarial trainingSSC networks · SSC (literature)performance: 10%-15% boostImproved
use early fusion by concatenating modalities instead of single inputSSC networks · SSC (literature)performanceMixed

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