AI Paper Atlas

Research map / Point clouds / Point cloud classification

Utonia: Toward One Encoder for All Point Clouds

Yujia Zhang, Xiaoyang Wu, Yunhan Yang, Xianzhe Fan, Han Li, Yuechen Zhang, Zehao Huang, Naiyan Wang, Hengshuang Zhao

ICML 2026 · Point clouds › Point cloud classification · approach: Self-supervised point cloud pretraining · 0 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 RoPE on granularity-aligned coordinates (ablation)Utonia multi-data · ScanNet200 / Waymo / PartNetEmIoU: 33.5 -> 34.9; 56.6 -> 59.2; 43.5 -> 44.6Improved
add RoPE in single-domain training (ablation)Utonia single-data · ScanNet200 / PartNetEmIoU: 34.4 -> 34.0; 44.1 -> 42.4Hurt
add RoPE in single-domain training (ablation)Utonia single-data · WaymomIoU: 60.5 -> 62.1Improved
vary RoPE base from 1 to 1000 (ablation)Utonia · ScanNet200 / Waymo / PartNetEmIoU: 34.0, 34.9, 34.1, 34.9 on ScanNet200No clear effect
add modality blinding at data loading (ablation)Utonia · ScanNet200 w/o colormIoU: 9.3 -> 31.7Improved
drop modality on masked views instead of at loading (ablation)Utonia · ScanNet200 w/o colormIoU: 31.7 -> 10.1Hurt
use full-strength object scale augmentation (vs less scale aug.) (ablation)Utonia · ScanObjectNN(H)mIoU: 64.5 -> 66.9Improved
use strong object rotation augmentation (vs less rot. aug.) (ablation)Utonia · ScanObjectNN(H)mIoU: 63.0 -> 66.9Improved
use strong rotation augmentation for objects (vs less rot. aug.) (ablation)Utonia · PartNetEmIoU: 47.5 -> 44.6Hurt
add temporal frame augmentation (ablation)Utonia · PartNetE / WaymomIoU: 40.0 -> 44.6; 58.8 -> 59.2Improved
add RoPE coordinate augmentation (ablation)Utonia · WaymomIoU: 60.1 -> 59.2No clear effect
scale model from 38M to 137M with 83k data (ablation)Utonia · Waymo linear probingmIoU: 59.2 -> 62.6Improved

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