AI Paper Atlas

Research map / Point clouds / Shape and surface completion

PoinTr: Diverse Point Cloud Completion With Geometry-Aware Transformers

Xumin Yu, Yongming Rao, Ziyi Wang, Zuyan Liu, Jiwen Lu, Jie Zhou

ICCV 2021 · Point clouds › Shape and surface completion · approach: Structure-guided completion · 646 citations (OpenAlex)

Read the paper

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 query generator between encoder and decoder (ablation)PoinTr (vanilla transformer) · PCN datasetCD-l1 x1000: 9.43 -> 9.09Improved
add DGCNN feature extractor (ablation)PoinTr with query generator · PCN datasetCD-l1 x1000: 9.09 -> 8.69Improved
add geometry-aware block to all transformer blocks (ablation)PoinTr with query generator and DGCNN · PCN datasetCD-l1 x1000: 8.69 -> 8.44Improved
restrict geometry-aware block to the first transformer block only (instead of all) (ablation)PoinTr · PCN datasetCD-l1 x1000: 8.44 -> 8.38Improved
add query generator (ablation)PoinTr (vanilla transformer) · PCN datasetF-Score@1%: 67.82 -> 0.713Improved
add DGCNN feature extractor (ablation)PoinTr with query generator · PCN datasetF-Score@1%: 0.713 -> 0.736Improved
add geometry-aware block to first block only (ablation)PoinTr with query generator and DGCNN · PCN datasetF-Score@1%: 0.736 -> 0.745Improved
use PoinTr instead of GRNetcompletion network · ShapeNet-55CD-l2 avg x1000: 1.97 -> 1.09Improved
use PoinTr instead of GRNetcompletion network · ShapeNet-55F-Score@1%: 0.238 -> 0.464Improved
evaluate on unseen instead of seen categoriesPoinTr · ShapeNet-34CD-l2 avg x1000: 1.23 -> 2.05Hurt
use PoinTr instead of CRNcompletion network · PCN datasetCD-l1 x1000: 8.51 -> 8.38Improved
use PoinTr instead of GRNetcompletion network · KITTI LiDAR carsFidelity: 0.816 -> 0.000Improved

Later papers that compared with this method

27 later accepted papers report PoinTr as a baseline in their results tables; 23 of them report a better number on most of the results they share with it.

19 more not listed.

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 4 of 4 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