期刊
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
卷 68, 期 1, 页码 38-48出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TIM.2018.2840598
关键词
3-D object recognition; convolutional neural network (CNN); deep learning (DL); network binarization; volumetric representation
资金
- National Natural Science Foundation of China [61403265, 61602499, 61471371]
- Science and Technology Plan of Sichuan Province [2015SZ0226]
- National Postdoctoral Program for Innovative Talents [BX201600172]
- China Postdoctoral Science Foundation
To address the high computational and memory cost in 3-D volumetric convolutional neural networks (CNNs), we propose an approach to train binary volumetric CNNs for 3-D object recognition. Our method is specifically designed for 3-D data, in which it transforms the inputs and weights in convolutional/fully connected layers to binary values, which can potentially accelerate the networks by efficient bitwise operations. Two loss calculation methods are designed to solve the accuracy decrease problem when the weights in the last layer are binarized. Four binary volumetric CNNs are obtained from their corresponding floating-point networks using our approach. Evaluations on three public datasets from different domains (Computer Aided Design (CAD), light detection and ranging (LiDAR), and RGB-D) show that our binary volumetric CNNs can achieve a comparable recognition performance as their floating-point counterparts but consume less computational and memory resources.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据