AI Paper Atlas

Research map / 3D vision / Face and shape

Weakly-supervised 3D Shape Completion in the Wild

Jiayuan Gu, Wei-Chiu Ma, Sivabalan Manivasagam, Wenyuan Zeng, Zihao Wang, Yuwen Xiong, Hao Su, Raquel Urtasun

ECCV 2020 · 3D vision › Face and shape · approach: Primitive-based shape abstraction · 56 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 shape-projection-matching-observation term (ablation)Ours · ShapeNet Airplanechamfer distance x100: 2.80 -> 1.65Improved
add shape-projection-matching-observation term (ablation)Ours · ShapeNet Carchamfer distance x100: 2.96 -> 2.68Improved
add shape-projection-matching-observation term (ablation)Ours · ShapeNet Chairchamfer distance x100: 3.94 -> 3.33Improved
add hindsight loss (ablation)Ours · ShapeNet Airplanechamfer distance x100: 2.32 -> 1.65Improved
add hindsight loss (ablation)Ours · ShapeNet Carchamfer distance x100: 2.82 -> 2.68Improved
add hindsight loss (ablation)Ours · ShapeNet Chairchamfer distance x100: 3.80 -> 3.33Improved
add projection term (3D vehicle dataset) (ablation)Ours · 3D vehicle datasetprecision/coverage/chamfer distanceMixed
increase training views per instance from 2 to 8 (ablation)Ours · 3D vehicle datasetchamfer distance: 0.307 -> 0.242Improved
increase training views per instance from 2 to 4 (ablation)Ours · 3D vehicle datasetrotation error: 6.185 -> 4.208Improved
use weakly supervised canonical-pose method instead of DPCDPC · ShapeNet Airplane registrationpose accuracy: 74.17 -> 92.87Improved
use 2D-projected input/output loss variant for carsOurs · ShapeNet Car registrationpose accuracy: 84.75 -> 91.03Improved
evaluate against partial GT rather than symmetric complete GTOurs · 3D vehicle datasetchamfer distance: 0.255 -> 0.195No clear effect

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