AI Paper Atlas

Research map / Point clouds / Point cloud classification

FastPoint: Accelerating 3D Point Cloud Model Inference via Sample Point Distance Prediction

Donghyun Lee, Dawoon Jeong, Jae W. Lee, Hongil Yoon

ICCV 2025 · Point clouds › Point cloud classification · approach: Efficient point sampling architectures · 1 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
replace FPS with MDPS at inference (no retraining)PointVector-L / PointMetaBase-L · S3DIS, ScanNet, SemanticKITTImIoU: 71.33 -> 71.37 (PV-L S3DIS); 50.91 -> 50.79 (PV-L SemanticKITTI)No clear effect
replace FPS with random samplingPointVector-L · SemanticKITTImIoU: 50.91 -> 38.46Hurt
replace FPS with grid sampling at inference onlyPointVector-L · ScanNetmIoU: 70.70 -> 69.61Hurt
add Redundancy-Free Neighbor Search on top of MDPS (ablation)FastPoint · ScanNet, PointMetaBase-Lend-to-end speedup: 2.89x -> 3.38xImproved
increase number of MDPS segments from 1 to 6 (ablation)MDPS · ScanNet, PointVector-LmIoU: 70.34 -> 70.63Improved
increase number of MDPS segments from 1 to 6 (ablation)MDPS · ScanNet, PointVector-Lsampling time (ms): 18.62 -> 22.06Hurt
reduce initial FPS ratio p from 0.2 to 0.05 (ablation)FastPoint · ScanNet, PointMetaBase-Lend-to-end speedup: 2.70x -> 3.82xImproved
reduce initial FPS ratio p from 0.2 to 0.05 (ablation)FastPoint · ScanNet, PointMetaBase-LmIoU: 70.95 -> 70.83Hurt
use MLP instead of power function for curve estimator (ablation)MDPS · SemanticKITTIMAPE (%): 6.6118 -> 1.9318Improved
apply estimator trained on SemanticKITTI to ScanNet (ablation)FastPoint · ScanNet, PointMetaBase-Lestimation error (%): 0.77 -> 2.24Hurt
apply data augmentation to validation data (ablation)FastPoint · ScanNet, PointMetaBase-Lsampling quality (%): 99.35 -> 99.10No clear effect
combine FastPoint with QuickFPS for initial FPS iterationsFastPoint · S3DIS, ScanNet, SemanticKITTIend-to-end speedup vs FPS: 2.55x -> 2.76xImproved

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