AI Paper Atlas

Research map / Point clouds / Shape and surface completion

Learning Gradient Fields for Shape Generation

Ruojin Cai, Guandao Yang, Hadar Averbuch-Elor, Zekun Hao, Serge Belongie, Noah Snavely, Bharath Hariharan

ECCV 2020 · Point clouds › Shape and surface completion · approach: Point cloud diffusion generation · 245 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
train with multiple noise levels instead of a single level 0.1 (ablation)gradient field model · Airplane auto-encodingCD (x1e4): 2.545 -> 0.993Improved
train with multiple noise levels instead of a single level 0.05 (ablation)gradient field model · Airplane auto-encodingCD (x1e4): 1.573 -> 0.993Improved
train with multiple noise levels instead of a single level 0.01 (ablation)gradient field model · Airplane auto-encodingCD (x1e4): 1009.357 -> 0.993Improved
initialize prior with Gaussian or fixed point instead of uniform (ablation)gradient field model · Airplane auto-encodingCD (x1e4): 0.993 -> 0.996No clear effect
add shortcuts, CBN latent input and decoder concat on top of (a) (ablation)architecture · Airplane auto-encodingCD (x1e4): 1.234 -> 0.987Improved
add shortcuts to CBN variant (ab vs a) (ablation)architecture · Airplane auto-encodingCD (x1e4): 1.234 -> 0.992Improved
use latent GAN instead of VAE for shape distribution (ablation)gradient field model · Airplane generationMMD-CD (x1e3): 1.909 -> 1.332Improved
use latent GAN instead of VAE for shape distribution (ablation)gradient field model · Chair generationMMD-CD (x1e3): 18.032 -> 14.818Improved
reduce input points M from 2048 to 128 (ablation)gradient field model · Airplane sparse auto-encoding (N=10K)CD (x1e4): 0.993 -> 1.688Hurt
reduce number of available points N from 10K to 1Kgradient field model · Airplane sparse auto-encoding (M=512)CD (x1e4): 0.999 -> 1.021No clear effect
use gradient field model instead of PointFlowgradient field model · Chair generationCOV-CD (%): 40.94 -> 46.37Improved
use gradient field model instead of PointFlowgradient field model · Airplane generation1-NNA-CD (%): 83.21 -> 85.06Hurt

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