A GPU-Based Framework for Generating Implicit Datasets of Voxelized Polygonal Models for the Training of 3D Convolutional Neural Networks

In this paper we present an efficient GPU-based framework to dynamically perform the voxelization of polygonal models for training 3D convolutional neural networks. It is designed to manage the dataset augmentation by using efficient geometric transformations and random vertex displacements in GPU....

Descripción completa

Detalles Bibliográficos
Autores: Ogáyar-Anguita, Carlos-Javier, Rueda-Ruiz, Antonio-Jesús, Segura-Sánchez, Rafael-Jesús, Díaz-Medina, Miguel, García-Fernández, Ángel-Luis
Tipo de recurso: artículo
Estado:Versión aceptada para publicación
Fecha de publicación:2020
País:España
Institución:Universidad de Jaén
Repositorio:RUJA. Repositorio Institucional de la Producción Científica de la Universidad de Jaén
OAI Identifier:oai:ruja.ujaen.es:10953/3304
Acceso en línea:https://doi.org/10.1109/ACCESS.2020.2965624
https://hdl.handle.net/10953/3304
Access Level:acceso abierto
Palabra clave:Voxelization
B-Rep
Boundary representation
Polygonal meshes
Convolutional neural network
3D-CNN
Geometric deep learning
004.92 - Computer graphics
004.8 - Artificial intelligence
Descripción
Sumario:In this paper we present an efficient GPU-based framework to dynamically perform the voxelization of polygonal models for training 3D convolutional neural networks. It is designed to manage the dataset augmentation by using efficient geometric transformations and random vertex displacements in GPU. With the proposed system, every voxelization is carried out on-the-fly for directly feeding the network. The computing performance with this approach is much better than with the standard method, that carries out every voxelization of each model separately and has much higher data processing overhead. The core voxelization algorithm works by using the standard rendering pipeline of the GPU. It generates binary voxels for both the interior and the surface of the models. Moreover, it is simple, concise and very compatible with commodity hardware, as it neither uses complex data structures nor needs vendor-specific hardware or additional dependencies. This framework dramatically reduces the input/output operations, and completely eliminates the storage footprint of the voxelization dataset, managing it as an implicit dataset.