· 2018
SO-Net: Self-Organizing Network for Point Cloud Analysis
- 984citations
- 2018year
Short summary
SO-Net, a novel deep learning architecture, uses a Self-Organizing Map (SOM) to process orderless point clouds, achieving state-of-the-art performance in tasks like classification and reconstruction while training significantly faster than existing methods.
AI-generated from the title and abstract; the full text is not read.
Key points
- SO-Net utilizes a Self-Organizing Map (SOM) to model spatial distribution and enable permutation-invariant processing of point clouds.
- The network performs hierarchical feature extraction on individual points and SOM nodes, representing the entire point cloud with a single feature vector.
- Receptive field size can be adjusted by tuning the k-nearest neighbor search between points and SOM nodes.
- SO-Net achieves competitive or superior performance in point cloud reconstruction, classification, object part segmentation, and shape retrieval.
- The architecture's parallelizability and simplicity result in significantly faster training times compared to existing networks.
AI-generated from the title and abstract; the full text is not read.
Abstract
This paper presents SO-Net, a permutation invariant architecture for deep learning with orderless point clouds. The SO-Net models the spatial distribution of point cloud by building a Self-Organizing Map (SOM). Based on the SOM, SO-Net performs hierarchical feature extraction on individual points and SOM nodes, and ultimately represents the input point cloud by a single feature vector. The receptive field of the network can be systematically adjusted by conducting point-to-node k nearest neighbor search. In recognition tasks such as point cloud reconstruction, classification, object part segmentation and shape retrieval, our proposed network demonstrates performance that is similar with or better than state-of-the-art approaches. In addition, the training speed is significantly faster than existing point cloud recognition networks because of the parallelizability and simplicity of the proposed architecture. Our code is available at the project website1.
The authors' abstract, as published at the source., 2018 · DOI ↗
Continue with a free account
Ask the paper: 3 free questions a day about this paper; save it, get its citation, new summaries every day for your field. Takeaways are Premium.
Continue free on the webSign in with Google or Apple; no card needed. You come back to this paper.
On your phone:
Field: Computational Mechanics
Computational MechanicsEngineering