Shape Representation Of 3D Point Clouds on ScanObjectNN

Metric: Overall Accuracy (higher is better)

LeaderboardDataset

Loading chart...

Results

Hide extra data

Sort:

#	Model↕	Overall Accuracy▼	Extra Data	Paper	Date↕	Code
1	OmniVec2	97.2	Yes	-	-	-
2	PointGST	96.18	Yes	Parameter-Efficient Fine-Tuning in Spectral Doma...	2024-10-10	Code
3	OmniVec	96.1	Yes	OmniVec: Learning robust representations with cr...	2023-11-07	-
4	GPSFormer	95.4	No	GPSFormer: A Global Perception and Local Structu...	2024-07-18	Code
5	ReCon++	95.25	Yes	ShapeLLM: Universal 3D Object Understanding for ...	2024-02-27	Code
6	AsymDSD-B* (no voting)	93.72	Yes	Asymmetric Dual Self-Distillation for 3D Self-Su...	2025-06-26	Code
7	PointGPT	93.4	Yes	PointGPT: Auto-regressively Generative Pre-train...	2023-05-19	Code
8	GPSFormer-elite	93.3	No	GPSFormer: A Global Perception and Local Structu...	2024-07-18	Code
9	Mamba3D	92.64	No	Mamba3D: Enhancing Local Features for 3D Point C...	2024-04-23	Code
10	Mamba3D (no voting)	91.81	No	Mamba3D: Enhancing Local Features for 3D Point C...	2024-04-23	Code
11	ULIP-2 + PointNeXt	91.5	Yes	ULIP-2: Towards Scalable Multimodal Pre-training...	2023-05-14	Code
12	ReCon	91.26	Yes	Contrast with Reconstruct: Contrastive 3D Repres...	2023-02-05	Code
13	ULIP-2 + PointNeXt (no voting)	90.8	Yes	ULIP-2: Towards Scalable Multimodal Pre-training...	2023-05-14	Code
14	ReCon (no voting)	90.63	Yes	Contrast with Reconstruct: Contrastive 3D Repres...	2023-02-05	Code
15	AsymDSD-S (no voting)	90.53	Yes	Asymmetric Dual Self-Distillation for 3D Self-Su...	2025-06-26	Code
16	DeLA	90.4	No	Decoupled Local Aggregation for Point Cloud Lear...	2023-08-31	Code
17	PCP-MAE	90.35	No	PCP-MAE: Learning to Predict Centers for Point M...	2024-08-16	Code
18	PointConT	90.3	No	Point Cloud Classification Using Content-based T...	2023-03-08	Code
19	Point-RAE (no voting)	90.28	Yes	Regress Before Construct: Regress Autoencoder fo...	2023-09-25	Code
20	Point-FEMAE	90.22	Yes	Towards Compact 3D Representations via Point Fea...	2023-12-17	Code
21	I2P-MAE (no voting)	90.11	Yes	Learning 3D Representations from 2D Pre-trained ...	2022-12-13	Code
22	ULIP + PointNeXt	89.7	Yes	ULIP: Learning a Unified Representation of Langu...	2022-12-10	Code
23	ReCon+PPT	89.52	No	Positional Prompt Tuning for Efficient 3D Repres...	2024-08-21	Code
24	3D-JEPA	89.52	Yes	3D-JEPA: A Joint Embedding Predictive Architectu...	2024-09-24	-
25	PointMLP∗ + JM3D	89.5	Yes	Beyond First Impressions: Integrating Joint Mult...	2023-08-06	Code
26	ULIP + PointMLP	89.4	Yes	ULIP: Learning a Unified Representation of Langu...	2022-12-10	Code
27	KPConvX-L	89.3	No	KPConvX: Modernizing Kernel Point Convolution wi...	2024-05-21	Code
28	P2P	89.3	Yes	P2P: Tuning Pre-trained Image Models for Point C...	2022-08-04	Code
29	ACT	89.17	Yes	Autoencoders as Cross-Modal Teachers: Can Pretra...	2022-12-16	Code
30	Ours	89	No	-	-	-
31	OTMae3D	89	Yes	-	-	Code
32	ULIP-2 + Point-BERT	89	Yes	ULIP-2: Towards Scalable Multimodal Pre-training...	2023-05-14	Code
33	PointNeXt+Local	88.6	No	Local Neighborhood Features for 3D Classification	2022-12-09	Code
34	SPoTr	88.6	No	Self-positioning Point-based Transformer for Poi...	2023-03-29	Code
35	IDPT	88.51	No	Instance-aware Dynamic Prompt Tuning for Pre-tra...	2023-04-14	Code
36	PointMLP+TAP	88.5	No	Take-A-Photo: 3D-to-2D Generative Pre-training o...	2023-07-27	Code
37	PointNeXt+GAM	88.4	No	-	-	-
38	PointNeXt+HyCoRe	88.3	No	Rethinking the compositionality of point clouds ...	2022-09-21	Code
39	ACT (no voting)	88.21	Yes	Autoencoders as Cross-Modal Teachers: Can Pretra...	2022-12-16	Code
40	PointNeXt	88.2	No	PointNeXt: Revisiting PointNet++ with Improved T...	2022-06-09	Code
41	PointConT (no voting)	88	No	Point Cloud Classification Using Content-based T...	2023-03-08	Code
42	PointVector-S	87.8	No	PointVector: A Vector Representation In Point Cl...	2022-05-21	Code
43	point2vec	87.5	Yes	Point2Vec for Self-Supervised Representation Lea...	2023-03-29	Code
44	PointStack	87.2	No	Advanced Feature Learning on Point Clouds using ...	2022-05-20	Code
45	Point-PN	87.1	No	Parameter is Not All You Need: Starting from Non...	2023-03-14	Code
46	PointCMT	86.7	No	Let Images Give You More:Point Cloud Cross-Modal...	2022-10-09	Code
47	Point-JEPA	86.6	Yes	Point-JEPA: A Joint Embedding Predictive Archite...	2024-04-25	Code
48	PointMLS	86.6	No	ModelNet-O: A Large-Scale Synthetic Dataset for ...	2024-01-16	Code
49	Point-M2AE	86.43	Yes	Point-M2AE: Multi-scale Masked Autoencoders for ...	2022-05-28	Code
50	ULIP + PointBERT	86.4	Yes	ULIP: Learning a Unified Representation of Langu...	2022-12-10	Code
51	DualMLP	86.4	No	-	-	Code
52	RepSurf-U (2x)	86	No	Surface Representation for Point Clouds	2022-05-11	Code
53	PointMLP	85.7	No	Rethinking Network Design and Local Geometry in ...	2022-02-15	Code
54	Point-LGMask	85.3	Yes	-	-	Code
55	Point-MAE	85.2	Yes	Masked Autoencoders for Point Cloud Self-supervi...	2022-03-13	Code
56	MVTN+SimpleView++	84.8	No	-	-	Code
57	DeltaConv	84.7	No	DeltaConv: Anisotropic Operators for Geometric D...	2021-11-16	Code
58	RepSurf-U	84.6	No	Surface Representation for Point Clouds	2022-05-11	Code
59	APP-Net	84.1	No	APP-Net: Auxiliary-point-based Push and Pull Ope...	2022-05-02	Code
60	PointMLP-elite	83.8	No	Rethinking Network Design and Local Geometry in ...	2022-02-15	Code
61	PointNet++ + SageMix	83.7	No	SageMix: Saliency-Guided Mixup for Point Clouds	2022-10-13	Code
62	DGCNN + SageMix	83.6	No	SageMix: Saliency-Guided Mixup for Point Clouds	2022-10-13	Code
63	Point-TnT	83.5	No	Points to Patches: Enabling the Use of Self-Atte...	2022-04-08	Code
64	Point-BERT	83.1	Yes	Point-BERT: Pre-training 3D Point Cloud Transfor...	2021-11-29	Code
65	MVTN	82.8	No	MVTN: Multi-View Transformation Network for 3D S...	2020-11-26	Code
66	PRA-Net	82.1	No	PRA-Net: Point Relation-Aware Network for 3D Poi...	2021-12-09	Code
67	DynamicScale	82	No	-	-	Code
68	PatchAugment	81	No	-	-	Code
69	GBNet	80.5	No	Geometric Back-projection Network for Point Clou...	2019-11-28	Code
70	SimpleView	80.5	No	-	-	Code
71	DRNet	80.3	No	Dense-Resolution Network for Point Cloud Classif...	2020-05-14	Code
72	PointCNN	78.5	No	PointCNN: Convolution On $\mathcal{X}$-Transform...	2018-01-23	Code
73	DGCNN	78.1	No	Dynamic Graph CNN for Learning on Point Clouds	2018-01-24	Code
74	PointNet++	77.9	No	PointNet++: Deep Hierarchical Feature Learning o...	2017-06-07	Code
75	SpiderCNN	73.7	No	SpiderCNN: Deep Learning on Point Sets with Para...	2018-03-30	Code
76	PointNet	68.2	No	PointNet: Deep Learning on Point Sets for 3D Cla...	2016-12-02	Code

#1OmniVec2
97.2
Overall Accuracy· Extra Data
No paper
#2PointGSTSOTA
96.18
Overall Accuracy· Extra Data· 2024-10-10
Parameter-Efficient Fine-Tuning in Spectral Domain for Point Cloud Learning Code
#3OmniVecSOTA
96.1
Overall Accuracy· Extra Data· 2023-11-07
OmniVec: Learning robust representations with cross modal sharing
#4GPSFormer
95.4
Overall Accuracy· 2024-07-18
GPSFormer: A Global Perception and Local Structure Fitting-based Transformer for Point Cloud Understanding Code
#5ReCon++
95.25
Overall Accuracy· Extra Data· 2024-02-27
ShapeLLM: Universal 3D Object Understanding for Embodied Interaction Code
#6AsymDSD-B* (no voting)
93.72
Overall Accuracy· Extra Data· 2025-06-26
Asymmetric Dual Self-Distillation for 3D Self-Supervised Representation Learning Code
#7PointGPTSOTA
93.4
Overall Accuracy· Extra Data· 2023-05-19
PointGPT: Auto-regressively Generative Pre-training from Point Clouds Code
#8GPSFormer-elite
93.3
Overall Accuracy· 2024-07-18
GPSFormer: A Global Perception and Local Structure Fitting-based Transformer for Point Cloud Understanding Code
#9Mamba3D
92.64
Overall Accuracy· 2024-04-23
Mamba3D: Enhancing Local Features for 3D Point Cloud Analysis via State Space Model Code
#10Mamba3D (no voting)
91.81
Overall Accuracy· 2024-04-23
Mamba3D: Enhancing Local Features for 3D Point Cloud Analysis via State Space Model Code
#11ULIP-2 + PointNeXtSOTA
91.5
Overall Accuracy· Extra Data· 2023-05-14
ULIP-2: Towards Scalable Multimodal Pre-training for 3D Understanding Code
#12ReConSOTA
91.26
Overall Accuracy· Extra Data· 2023-02-05
Contrast with Reconstruct: Contrastive 3D Representation Learning Guided by Generative Pretraining Code
#13ULIP-2 + PointNeXt (no voting)
90.8
Overall Accuracy· Extra Data· 2023-05-14
ULIP-2: Towards Scalable Multimodal Pre-training for 3D Understanding Code
#14ReCon (no voting)
90.63
Overall Accuracy· Extra Data· 2023-02-05
Contrast with Reconstruct: Contrastive 3D Representation Learning Guided by Generative Pretraining Code
#15AsymDSD-S (no voting)
90.53
Overall Accuracy· Extra Data· 2025-06-26
Asymmetric Dual Self-Distillation for 3D Self-Supervised Representation Learning Code
#16DeLA
90.4
Overall Accuracy· 2023-08-31
Decoupled Local Aggregation for Point Cloud Learning Code
#17PCP-MAE
90.35
Overall Accuracy· 2024-08-16
PCP-MAE: Learning to Predict Centers for Point Masked Autoencoders Code
#18PointConT
90.3
Overall Accuracy· 2023-03-08
Point Cloud Classification Using Content-based Transformer via Clustering in Feature Space Code
#19Point-RAE (no voting)
90.28
Overall Accuracy· Extra Data· 2023-09-25
Regress Before Construct: Regress Autoencoder for Point Cloud Self-supervised Learning Code
#20Point-FEMAE
90.22
Overall Accuracy· Extra Data· 2023-12-17
Towards Compact 3D Representations via Point Feature Enhancement Masked Autoencoders Code
#21I2P-MAE (no voting)SOTA
90.11
Overall Accuracy· Extra Data· 2022-12-13
Learning 3D Representations from 2D Pre-trained Models via Image-to-Point Masked Autoencoders Code
#22ULIP + PointNeXtSOTA
89.7
Overall Accuracy· Extra Data· 2022-12-10
ULIP: Learning a Unified Representation of Language, Images, and Point Clouds for 3D Understanding Code
#23ReCon+PPT
89.52
Overall Accuracy· 2024-08-21
Positional Prompt Tuning for Efficient 3D Representation Learning Code
#243D-JEPA
89.52
Overall Accuracy· Extra Data· 2024-09-24
3D-JEPA: A Joint Embedding Predictive Architecture for 3D Self-Supervised Representation Learning
#25PointMLP∗ + JM3D
89.5
Overall Accuracy· Extra Data· 2023-08-06
Beyond First Impressions: Integrating Joint Multi-modal Cues for Comprehensive 3D Representation Code
#26ULIP + PointMLP
89.4
Overall Accuracy· Extra Data· 2022-12-10
ULIP: Learning a Unified Representation of Language, Images, and Point Clouds for 3D Understanding Code
#27KPConvX-L
89.3
Overall Accuracy· 2024-05-21
KPConvX: Modernizing Kernel Point Convolution with Kernel Attention Code
#28P2PSOTA
89.3
Overall Accuracy· Extra Data· 2022-08-04
P2P: Tuning Pre-trained Image Models for Point Cloud Analysis with Point-to-Pixel Prompting Code
#29ACT
89.17
Overall Accuracy· Extra Data· 2022-12-16
Autoencoders as Cross-Modal Teachers: Can Pretrained 2D Image Transformers Help 3D Representation Learning?Code
#30Ours
89
Overall Accuracy
No paper
#31OTMae3D
89
Overall Accuracy· Extra Data
No paperCode
#32ULIP-2 + Point-BERT
89
Overall Accuracy· Extra Data· 2023-05-14
ULIP-2: Towards Scalable Multimodal Pre-training for 3D Understanding Code
#33PointNeXt+Local
88.6
Overall Accuracy· 2022-12-09
Local Neighborhood Features for 3D Classification Code
#34SPoTr
88.6
Overall Accuracy· 2023-03-29
Self-positioning Point-based Transformer for Point Cloud Understanding Code
#35IDPT
88.51
Overall Accuracy· 2023-04-14
Instance-aware Dynamic Prompt Tuning for Pre-trained Point Cloud Models Code
#36PointMLP+TAP
88.5
Overall Accuracy· 2023-07-27
Take-A-Photo: 3D-to-2D Generative Pre-training of Point Cloud Models Code
#37PointNeXt+GAM
88.4
Overall Accuracy
No paper
#38PointNeXt+HyCoRe
88.3
Overall Accuracy· 2022-09-21
Rethinking the compositionality of point clouds through regularization in the hyperbolic space Code
#39ACT (no voting)
88.21
Overall Accuracy· Extra Data· 2022-12-16
Autoencoders as Cross-Modal Teachers: Can Pretrained 2D Image Transformers Help 3D Representation Learning?Code
#40PointNeXtSOTA
88.2
Overall Accuracy· 2022-06-09
PointNeXt: Revisiting PointNet++ with Improved Training and Scaling Strategies Code
#41PointConT (no voting)
88
Overall Accuracy· 2023-03-08
Point Cloud Classification Using Content-based Transformer via Clustering in Feature Space Code
#42PointVector-SSOTA
87.8
Overall Accuracy· 2022-05-21
PointVector: A Vector Representation In Point Cloud Analysis Code
#43point2vec
87.5
Overall Accuracy· Extra Data· 2023-03-29
Point2Vec for Self-Supervised Representation Learning on Point Clouds Code
#44PointStackSOTA
87.2
Overall Accuracy· 2022-05-20
Advanced Feature Learning on Point Clouds using Multi-resolution Features and Learnable Pooling Code
#45Point-PN
87.1
Overall Accuracy· 2023-03-14
Parameter is Not All You Need: Starting from Non-Parametric Networks for 3D Point Cloud Analysis Code
#46PointCMT
86.7
Overall Accuracy· 2022-10-09
Let Images Give You More:Point Cloud Cross-Modal Training for Shape Analysis Code
#47Point-JEPA
86.6
Overall Accuracy· Extra Data· 2024-04-25
Point-JEPA: A Joint Embedding Predictive Architecture for Self-Supervised Learning on Point Cloud Code
#48PointMLS
86.6
Overall Accuracy· 2024-01-16
ModelNet-O: A Large-Scale Synthetic Dataset for Occlusion-Aware Point Cloud Classification Code
#49Point-M2AE
86.43
Overall Accuracy· Extra Data· 2022-05-28
Point-M2AE: Multi-scale Masked Autoencoders for Hierarchical Point Cloud Pre-training Code
#50ULIP + PointBERT
86.4
Overall Accuracy· Extra Data· 2022-12-10
ULIP: Learning a Unified Representation of Language, Images, and Point Clouds for 3D Understanding Code
#51DualMLP
86.4
Overall Accuracy
No paperCode
#52RepSurf-U (2x)SOTA
86
Overall Accuracy· 2022-05-11
Surface Representation for Point Clouds Code
#53PointMLPSOTA
85.7
Overall Accuracy· 2022-02-15
Rethinking Network Design and Local Geometry in Point Cloud: A Simple Residual MLP Framework Code
#54Point-LGMask
85.3
Overall Accuracy· Extra Data
No paperCode
#55Point-MAE
85.2
Overall Accuracy· Extra Data· 2022-03-13
Masked Autoencoders for Point Cloud Self-supervised Learning Code
#56MVTN+SimpleView++
84.8
Overall Accuracy
No paperCode
#57DeltaConvSOTA
84.7
Overall Accuracy· 2021-11-16
DeltaConv: Anisotropic Operators for Geometric Deep Learning on Point Clouds Code
#58RepSurf-U
84.6
Overall Accuracy· 2022-05-11
Surface Representation for Point Clouds Code
#59APP-Net
84.1
Overall Accuracy· 2022-05-02
APP-Net: Auxiliary-point-based Push and Pull Operations for Efficient Point Cloud Classification Code
#60PointMLP-elite
83.8
Overall Accuracy· 2022-02-15
Rethinking Network Design and Local Geometry in Point Cloud: A Simple Residual MLP Framework Code
#61PointNet++ + SageMix
83.7
Overall Accuracy· 2022-10-13
SageMix: Saliency-Guided Mixup for Point Clouds Code
#62DGCNN + SageMix
83.6
Overall Accuracy· 2022-10-13
SageMix: Saliency-Guided Mixup for Point Clouds Code
#63Point-TnT
83.5
Overall Accuracy· 2022-04-08
Points to Patches: Enabling the Use of Self-Attention for 3D Shape Recognition Code
#64Point-BERT
83.1
Overall Accuracy· Extra Data· 2021-11-29
Point-BERT: Pre-training 3D Point Cloud Transformers with Masked Point Modeling Code
#65MVTNSOTA
82.8
Overall Accuracy· 2020-11-26
MVTN: Multi-View Transformation Network for 3D Shape Recognition Code
#66PRA-Net
82.1
Overall Accuracy· 2021-12-09
PRA-Net: Point Relation-Aware Network for 3D Point Cloud Analysis Code
#67DynamicScale
82
Overall Accuracy
No paperCode
#68PatchAugment
81
Overall Accuracy
No paperCode
#69GBNetSOTA
80.5
Overall Accuracy· 2019-11-28
Geometric Back-projection Network for Point Cloud Classification Code
#70SimpleView
80.5
Overall Accuracy
No paperCode
#71DRNet
80.3
Overall Accuracy· 2020-05-14
Dense-Resolution Network for Point Cloud Classification and Segmentation Code
#72PointCNNSOTA
78.5
Overall Accuracy· 2018-01-23
PointCNN: Convolution On $\mathcal{X}$-Transformed Points Code
#73DGCNN
78.1
Overall Accuracy· 2018-01-24
Dynamic Graph CNN for Learning on Point Clouds Code
#74PointNet++SOTA
77.9
Overall Accuracy· 2017-06-07
PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space Code
#75SpiderCNN
73.7
Overall Accuracy· 2018-03-30
SpiderCNN: Deep Learning on Point Sets with Parameterized Convolutional Filters Code
#76PointNetSOTA
68.2
Overall Accuracy· 2016-12-02
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation Code