AI Paper Atlas

Research map / 3D vision / Face and shape

LAMP: Data-Efficient Linear Affine Weight-Space Models for Parameter-Controlled 3D Shape Generation and Extrapolation

Ghadi Nehme, Yanxia Zhang, Dule Shu, Matthew Klenk, Faez Ahmed

ICML 2026 · 3D vision › Face and shape · approach: 3D shape generative modeling

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
train exemplar SDF decoders from a shared initialization (ablation)LAMP · DrivAerNet++ single-parameter sweep, 100 exemplarsR^2: -37.77 -> 0.838Improved
train exemplar decoders with different random seeds (shared init kept) (ablation)LAMP · DrivAerNet++ single-parameter sweepR^2: 0.838 -> 0.773Hurt
increase number of exemplars from 10 to 100 (ablation)LAMP · DrivAerNet++R^2: -7.289 -> 0.838Improved
increase number of exemplars from 100 to 1000 (ablation)LAMP · DrivAerNet++R^2: 0.838 -> 0.862No clear effect
mix aligned SDF weights affinely instead of DNI weight predictionLAMP · DrivAerNet++ large-range extrapolation, 4 parametersR^2: -5.768 -> 0.867Improved
LAMP instead of AE-LPA with 1000 samplesLAMP with 100 samples · DrivAerNet++ interpolationIoU (%): 92.63 -> 97.24Improved
LAMP instead of DNI for interpolation within dataset rangeLAMP · DrivAerNet++ interpolationIoU (%): 97.12 -> 97.24No clear effect
use a pretrained OpenShape embedding for the surrogate instead of random PointNet (ablation)mesh-based surrogate · DrivAerNet++ parameter predictionR^2Hurt
filter generations with linearity-mismatch threshold 0.01LAMP · human-annotated valid/invalid meshesROC AUC: ROC AUC 0.989, PR AUC 0.990Improved
add performance (drag) target to LAMP mixingLAMP · DrivAerNet++ drag optimizationdrag decay MAE (%): 5.2 (AE-LPA) -> 2.7Improved

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