AI Paper Atlas

Research map / Point clouds / Shape and surface completion

LION: Latent Point Diffusion Models for 3D Shape Generation

NeurIPS 2022 · Point clouds › Shape and surface completion · approach: Point cloud diffusion generation · 190 citations (OpenAlex)

Read the paperCheck your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's title only (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
run diffusion in hierarchical VAE latent space (LION) instead of directly on point clouds (PVD)point cloud DDM · ShapeNet Airplane (PointFlow)1-NNA CD: 73.82 -> 67.41Helped
latent hierarchical DDM instead of direct point DDMpoint cloud DDM · ShapeNet-vol Airplane1-NNA CD: 66.46 -> 53.47Helped
latent hierarchical DDM instead of direct point DDMunconditional 13-class generation · ShapeNet-vol 13 classes1-NNA EMD: 57.85 -> 48.95Helped
use fine-tuned encoders plus latent diffuse-denoise for voxel-guided synthesisLION vs PVD/DPM · ShapeNet airplane voxel/outlier inputsreconstruction CD/EMDHelped
use 25-step DDIM sampling instead of 1000-step DDPMLION · ShapeNetsampling time: 27.12 s -> 0.89 sHelped
spherical instead of linear interpolation in DDM prior (ablation)LION interpolation · ShapeNetinterpolation quality (qualitative)Helped

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