AI Paper Atlas

Research map / Point clouds / Shape and surface completion

Snowflake Point Deconvolution for Point Cloud Completion and Generation with Skip-Transformer

Xiang Peng, Xin Wen, Yu-Shen Liu, Yan‐Pei Cao, Pengfei Wan, Wen Zheng, Zhizhong Han

TPAMI 2022 · Point clouds › Shape and surface completion · approach: Structure-guided completion · 108 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
remove whole skip-transformer (no feature connection between SPD layers) (ablation)SnowflakeNet · Completion3D 4 categoriesCD x10^4: 9.39 -> 8.48Improved
remove transformer mechanism from skip-transformer, add features directly (ablation)SnowflakeNet · Completion3D 4 categoriesCD x10^4: 9.30 -> 8.48Improved
use skip-transformer instead of self-attention (ablation)SnowflakeNet · Completion3D 4 categoriesCD x10^4: 8.89 -> 8.48Improved
use point-wise splitting instead of folding-based feature expansion (ablation)SnowflakeNet · Completion3D 4 categoriesCD x10^4: 8.80 -> 8.48Improved
add preservation loss (ablation)SnowflakeNet · Completion3D 4 categoriesCD x10^4: 8.50 -> 8.48Improved
use SPD decoder on PCN encoder instead of PCN decoder (ablation)PCN · Completion3D 4 categoriesCD x10^4: 13.30 -> 8.93Improved
use three splitting steps instead of baseline two-stage design (ablation)SnowflakeNet · Completion3D 4 categoriesCD x10^4: 8.48 -> 8.66Hurt
use one-step splitting instead of baseline (ablation)SnowflakeNet · Completion3D 4 categoriesCD x10^4: 8.48 -> 9.53Hurt
increase k-NN neighbors from 4 to 32 in skip-transformer (ablation)SnowflakeNet · Completion3D 4 categoriesCD x10^4: 8.69 -> 8.37Improved
train with 0% Gaussian noise augmentation (ablation)SnowflakeNet · Completion3D noise-free inputCD x10^4: 8.48 -> 8.81Hurt
increase test noise to 2.0% (ablation)SnowflakeNet · Completion3D noisy inputCD x10^4: 8.81 -> 10.56Hurt
SnowflakeNet vs PMP-Net++ on PCNPMP-Net++ · PCNL1 CD x10^3: 7.56 -> 7.21Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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