AI Paper Atlas

Research map / Point clouds / Registration and matching

BUFFER-X: Towards Zero-Shot Point Cloud Registration in Diverse Scenes

Minkyun Seo, Hyungtae Lim, Kanghee Lee, Luca Carlone, Jaesik Park

ICCV 2025 · Point clouds › Registration and matching · approach: Learned feature-based registration · 18 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
add oracle tuning of voxel size and radiusBUFFER · KITTI (zero-shot)success rate (%): 0.00 -> 99.46Improved
add oracle tuning of voxel size and radiusGeoTransformer · KITTI (zero-shot)success rate (%): N/A -> 0.36Hurt
add scale alignment on top of oracle tuningGeoTransformer · KITTI (zero-shot)success rate (%): 0.36 -> 92.43Improved
add scale alignment on top of oracle tuningPredator · KITTI (zero-shot)success rate (%): 0.00 -> 77.29Improved
add scale alignment on top of oracle tuningFCGF · KITTI (zero-shot)success rate (%): 90.27 -> 94.41Improved
add scale alignment on top of oracle tuningBUFFER · KITTI (zero-shot)success rate (%): 99.46 -> 99.46No clear effect
add middle scale to local scale (ablation)BUFFER-X · 3DMatchsuccess rate (%): 84.06 -> 94.31Improved
add global scale to middle scale (ablation)BUFFER-X · 3DMatchsuccess rate (%): 93.38 -> 94.62Improved
add local scale to middle+global (ablation)BUFFER-X · 3DMatchsuccess rate (%): 94.62 -> 95.58Improved
add local scale to middle+global (ablation)BUFFER-X · 3DMatchspeed (Hz): 2.33 -> 1.81Hurt
add middle and global scales to local only (ablation)BUFFER-X · 3DMatchspeed (Hz): 5.61 -> 1.81Hurt
use multi-scale instead of only middle radius r_m (ablation)BUFFER-X · ScanNet++Fsuccess rate (%): 99.60 -> 99.90Improved

Later papers that compared with this method

1 later accepted paper reports BUFFER-X as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 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