Papers with Code · CC BY-SA 4.0

ScanObjectNN

아카이브 평가 77건 · 마지막 적재/변경 2026-10-30

평가 기록

과거 수집 기록이며 최신 순위가 아닙니다. 데이터셋 버전·split·실험 조건과 지표 방향이 검증되지 않은 결과는 직접 비교할 수 없습니다. 날짜는 논문 날짜이며 최신 평가일을 뜻하지 않습니다.

표가 화면보다 넓으면 좌우로 스크롤하여 논문과 구현 링크를 확인하세요.

원본 기록 순서 · 순위 미부여
모델	원본 지표·값	논문	구현
OmniVec2	Overall Accuracy 97.2	OmniVec2 - A Novel Transformer based Network for Large Scale Multimodal and Multitask Learning 2024-01-01	연결 없음
PointGST	Overall Accuracy 96.18 OBJ-BG (OA) 99.48 OBJ-ONLY (OA) 97.76	Parameter-Efficient Fine-Tuning in Spectral Domain for Point Cloud Learning 2024-10-10	코드 1
OmniVec	Overall Accuracy 96.1	OmniVec: Learning robust representations with cross modal sharing 2023-11-07	연결 없음
GPSFormer	Overall Accuracy 95.4 Mean Accuracy 93.8 Number of params 2.36M FLOPs 0.7G	GPSFormer: A Global Perception and Local Structure Fitting-based Transformer for Point Cloud Understanding 2024-07-18	코드 1
ReCon++	Overall Accuracy 95.25 OBJ-BG (OA) 98.80 OBJ-ONLY (OA) 97.59	ShapeLLM: Universal 3D Object Understanding for Embodied Interaction 2024-02-27	코드 1 · 코드 2 · 코드 3
AsymDSD-B* (no voting)	Overall Accuracy 93.72 OBJ-BG (OA) 96.73 OBJ-ONLY (OA) 94.32	Asymmetric Dual Self-Distillation for 3D Self-Supervised Representation Learning 2025-06-26	코드 1
PointGPT	Overall Accuracy 93.4 OBJ-BG (OA) 97.2 OBJ-ONLY (OA) 96.6	PointGPT: Auto-regressively Generative Pre-training from Point Clouds 2023-05-19	코드 1
GPSFormer-elite	Overall Accuracy 93.30 Mean Accuracy 92.51 Number of params 0.68M	GPSFormer: A Global Perception and Local Structure Fitting-based Transformer for Point Cloud Understanding 2024-07-18	코드 1
Mamba3D	Overall Accuracy 92.64 Number of params 16.9M FLOPs 3.9G OBJ-BG (OA) 94.49 OBJ-ONLY (OA) 92.43	Mamba3D: Enhancing Local Features for 3D Point Cloud Analysis via State Space Model 2024-04-23	코드 1
Mamba3D (no voting)	Overall Accuracy 91.81 Number of params 16.9M FLOPs 3.9G OBJ-BG (OA) 92.94 OBJ-ONLY (OA) 92.08	Mamba3D: Enhancing Local Features for 3D Point Cloud Analysis via State Space Model 2024-04-23	코드 1
ULIP-2 + PointNeXt	Overall Accuracy 91.5 Mean Accuracy 91.2 Number of params 1.4M	ULIP-2: Towards Scalable Multimodal Pre-training for 3D Understanding 2023-05-14	코드 1
ReCon	Overall Accuracy 91.26 OBJ-BG (OA) 95.35 OBJ-ONLY (OA) 93.80	Contrast with Reconstruct: Contrastive 3D Representation Learning Guided by Generative Pretraining 2023-02-05	코드 1 · 코드 2 · 코드 3 · 코드 4 · 코드 5
ULIP-2 + PointNeXt (no voting)	Overall Accuracy 90.8 Mean Accuracy 90.3 Number of params 1.4M	ULIP-2: Towards Scalable Multimodal Pre-training for 3D Understanding 2023-05-14	코드 1
ReCon (no voting)	Overall Accuracy 90.63 OBJ-BG (OA) 95.18 OBJ-ONLY (OA) 93.29	Contrast with Reconstruct: Contrastive 3D Representation Learning Guided by Generative Pretraining 2023-02-05	코드 1 · 코드 2 · 코드 3 · 코드 4 · 코드 5
AsymDSD-S (no voting)	Overall Accuracy 90.53 OBJ-BG (OA) 94.32 OBJ-ONLY (OA) 91.91	Asymmetric Dual Self-Distillation for 3D Self-Supervised Representation Learning 2025-06-26	코드 1
DeLA	Overall Accuracy 90.4 Mean Accuracy 89.3 Number of params 5.3M FLOPs 1.5G	Decoupled Local Aggregation for Point Cloud Learning 2023-08-31	코드 1
PCP-MAE	Overall Accuracy 90.35 OBJ-BG (OA) 95.52 OBJ-ONLY (OA) 94.32	PCP-MAE: Learning to Predict Centers for Point Masked Autoencoders 2024-08-16	코드 1
PointConT	Overall Accuracy 90.3 Mean Accuracy 88.5	Point Cloud Classification Using Content-based Transformer via Clustering in Feature Space 2023-03-08	코드 1
Point-RAE (no voting)	Overall Accuracy 90.28 OBJ-BG (OA) 95.53 OBJ-ONLY (OA) 93.63	Regress Before Construct: Regress Autoencoder for Point Cloud Self-supervised Learning 2023-09-25	코드 1
Point-FEMAE	Overall Accuracy 90.22 OBJ-BG (OA) 95.18 OBJ-ONLY (OA) 93.29	Towards Compact 3D Representations via Point Feature Enhancement Masked Autoencoders 2023-12-17	코드 1
I2P-MAE (no voting)	Overall Accuracy 90.11 OBJ-BG (OA) 94.15 OBJ-ONLY (OA) 91.57	Learning 3D Representations from 2D Pre-trained Models via Image-to-Point Masked Autoencoders 2022-12-13	코드 1 · 코드 2
ULIP + PointNeXt	Overall Accuracy 89.7 Mean Accuracy 88.6 Number of params 1.4M	ULIP: Learning a Unified Representation of Language, Images, and Point Clouds for 3D Understanding 2022-12-10	코드 1
ReCon+PPT	Overall Accuracy 89.52 OBJ-BG (OA) 95.01 OBJ-ONLY (OA) 93.28	Positional Prompt Tuning for Efficient 3D Representation Learning 2024-08-21	코드 1
3D-JEPA	Overall Accuracy 89.52 OBJ-BG (OA) 93.63 OBJ-ONLY (OA) 94.49	3D-JEPA: A Joint Embedding Predictive Architecture for 3D Self-Supervised Representation Learning 2024-09-24	연결 없음
PointMLP∗ + JM3D	Overall Accuracy 89.5 Mean Accuracy 88.7	Beyond First Impressions: Integrating Joint Multi-modal Cues for Comprehensive 3D Representation 2023-08-06	코드 1
ULIP + PointMLP	Overall Accuracy 89.4 Mean Accuracy 88.5	ULIP: Learning a Unified Representation of Language, Images, and Point Clouds for 3D Understanding 2022-12-10	코드 1
KPConvX-L	Overall Accuracy 89.3 Mean Accuracy 88.1	KPConvX: Modernizing Kernel Point Convolution with Kernel Attention 2024-05-21	코드 1
P2P	Overall Accuracy 89.3 Number of params 195.8M	P2P: Tuning Pre-trained Image Models for Point Cloud Analysis with Point-to-Pixel Prompting 2022-08-04	코드 1
ACT	Overall Accuracy 89.17	Autoencoders as Cross-Modal Teachers: Can Pretrained 2D Image Transformers Help 3D Representation Learning? 2022-12-16	코드 1 · 코드 2 · 코드 3 · 코드 4
Ours	Overall Accuracy 89.0 Mean Accuracy 87.2	Beyond local patches: Preserving global–local interactions by enhancing self-attention via 3D point cloud tokenization 2024-11-01	연결 없음