AI Paper Atlas

Research map / Point clouds / Shape and surface completion

Learning 3D Shape Completion Under Weak Supervision

David Stutz, Andreas Geiger

IJCV 2018 · Point clouds › Shape and surface completion · approach: Unsupervised shape correspondence · 105 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
amortize maximum likelihood inference with a learned encoder (ML -> AML)shape completion · SN-clean low resIoU: 0.756 -> 0.779Improved
use probabilistic encoder with KL (AML) instead of deterministic dAML (ablation)AML · SN-clean low resIoU: 0.784 -> 0.779No clear effect
fuse multiple views (k=1 -> k=3) (ablation)AML · SN-clean low resIoU: 0.779 -> 0.809Improved
fuse k=5 views instead of k=3 (ablation)AML · SN-clean low resIoU: 0.809 -> 0.804No clear effect
fuse k=5 views with full supervision (single view -> k=5) (ablation)Dai17 · SN-clean low resIoU: 0.872 -> 0.924Improved
use skip connections (Sup -> Dai17)fully supervised baseline · SN-clean low resIoU: 0.841 -> 0.872Improved
train weakly supervised AML versus fully supervised Dai17 (SN-noisy)AML · SN-noisy low resIoU: 0.836 vs 0.771Mixed
increase resolution from low to mediumAML · SN-cleanIoU: 0.779 -> 0.788Improved
increase resolution from low to highAML · SN-cleanIoU: 0.779 -> 0.796Improved
use AML versus data-driven Eng16 on KITTIAML · KITTI low resCompleteness (m): 0.13 -> 0.12Improved
use full supervision Sup instead of AMLSup · KITTI low resCompleteness (m): 0.12 -> 0.091Improved
train a single model across all ten ModelNet categories (AML vs Dai17 gap)AML · ModelNet10IoU: 0.646 vs 0.509Mixed

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