Categories Companies Research

Loading the SOTA2 catalog…

Explore

Categories Companies Research

Contribute

Join SOTA2 Contact

SOTA2

© 2026 SOTA2

Build in Bellevue, USA

WizWand is now SOTA2 Research

Papers Benchmarks Datasets Tasks

Point Cloud Classification research benchmarks – Page 4 · SOTA2 Research

Point Cloud Classification

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

ScanObjectNN T25 (test)

89.4Accuracy

7

Feb 26, 2026

ScanObjectNN HARD (PB-T50-RS)

86.43Accuracy

7

Feb 26, 2026

Sydney Urban Objects

78.4Mean F1

7

Feb 26, 2026

81.42Accuracy

6

May 19, 2026

MVImageNet (test)

88.9Overall Accuracy

6

Feb 26, 2026

Objaverse PointLLM

47.5GPT-4 Accuracy

5

Feb 26, 2026

ScanObjectNN (SONN) (test)

1003D-Adv Robustness

4

Mar 20, 2026

Point Cloud Classification benchmark (test)

83Accuracy

4

Feb 26, 2026

ScanObjectNN S-PB_T50_RS

23.32Accuracy

3

Feb 27, 2026

ScanObjectNN S-OBJ_BG

30.46Accuracy

3

Feb 27, 2026

Rue Cassette (test)

97.27Facade IoU

3

Feb 26, 2026

Rue Madame (test)

98.22Facade IoU

3

Feb 26, 2026

ScanObjectNN 51 (test)

15.38Accuracy

1

Feb 26, 2026

ModelNet40 57 (test)

20.18Accuracy

1

Feb 26, 2026

ModelNet10 57 (test)

30.23Accuracy

1

Feb 26, 2026

ScanObjectNN S-PB_T50_RS hardest (test)

—

—Accuracy

0

Feb 26, 2026

ScanObjectNN S-OBJ_ONLY (test)

—

—Accuracy

0

Feb 26, 2026

ModelNet40 aligned (test)

—

—Accuracy

0

Feb 26, 2026

ModelNet-C Augmented (WolfMix) (test)

—

—OA

0

Feb 26, 2026

ScanObjectNN Object 1.0

—

—Primary metric

0

Feb 26, 2026

Page 4 of 4

Popular datasets

ModelNet ScanObjectNN ShapeNet Synthetic Point Clouds ModelNet-C

Follow for updates

Join SOTA2 Contact