ScanObjectNN

Overview edit

Variants: ScanObjectNN

Associated Benchmarks

This dataset is used in 3 benchmarks:

• 3D Point Cloud Classification - Metrics: Overall Accuracy, Mean Accuracy, OBJ-BG (OA), OBJ-ONLY (OA), FLOPs, Number of params
• 3D Point Cloud Linear Classification - Metrics: Overall Accuracy
• Training-free 3D Point Cloud Classification - Metrics: Accuracy (%), Parameters, Need 3D Data?

Recent Benchmark Submissions

Research Papers

Recent papers with results on this dataset:

• Asymmetric Dual Self-Distillation for 3D Self-Supervised Representation Learning (2025) -
• Point-GN: A Non-Parametric Network Using Gaussian Positional Encoding for Point Cloud Classification (2024) -
• Parameter-Efficient Fine-Tuning in Spectral Domain for Point Cloud Learning (2024) -
• 3D-JEPA: A Joint Embedding Predictive Architecture for 3D Self-Supervised Representation Learning (2024) -
• Positional Prompt Tuning for Efficient 3D Representation Learning (2024) -
• PCP-MAE: Learning to Predict Centers for Point Masked Autoencoders (2024) -
• GPSFormer: A Global Perception and Local Structure Fitting-based Transformer for Point Cloud Understanding (2024) -
• KPConvX: Modernizing Kernel Point Convolution with Kernel Attention (2024) -
• Point-JEPA: A Joint Embedding Predictive Architecture for Self-Supervised Learning on Point Cloud (2024) -
• Mamba3D: Enhancing Local Features for 3D Point Cloud Analysis via State Space Model (2024) -
• ShapeLLM: Universal 3D Object Understanding for Embodied Interaction (2024) -
• ModelNet-O: A Large-Scale Synthetic Dataset for Occlusion-Aware Point Cloud Classification (2024) -
• Towards Compact 3D Representations via Point Feature Enhancement Masked Autoencoders (2023) -
• OmniVec: Learning robust representations with cross modal sharing (2023) -
• Regress Before Construct: Regress Autoencoder for Point Cloud Self-supervised Learning (2023) -
• Decoupled Local Aggregation for Point Cloud Learning (2023) -
• Beyond First Impressions: Integrating Joint Multi-modal Cues for Comprehensive 3D Representation (2023) -
• Take-A-Photo: 3D-to-2D Generative Pre-training of Point Cloud Models (2023) -
• ExpPoint-MAE: Better interpretability and performance for self-supervised point cloud transformers (2023) -
• PointGPT: Auto-regressively Generative Pre-training from Point Clouds (2023) -