Fully Binarized Convolutional Neural Network for Accelerating Edge Vision Computing

机译：完全二值化卷积神经网络，用于加速边缘视觉计算

获取原文

获取原文并翻译 | 示例

页面导航

摘要
著录项
相似文献
相关主题

摘要

We propose BitFlow-Net, a method to simplify binary CNN model to almost no floating-point multiplication at inference time. Recently, a lot of variations of binarized networks were proposed trying to achieve high accuracy while replacing resource-consuming floating-point multiplication with bit operation. These methods usually require scaling factor and BatchNorm to achieve comparable accuracy as their full-precision counterparts. However, data flow have to be frequently converted between floating-point data and bit data due to the multiplication with scaling factor and in BatchNorm. Such conversion will cost extra resources and time when implemented on edge hardware. Motivated by that, we further explore and reveal some basic attributes of BNN based on previous works and propose a new method to simplify binary network. As a result, our model could inference with most of its data flow remains bit flow. Such a network architecture will greatly reduce the design complexity when implemented on ASIC or FPGA. Our method performs no accuracy degradation on ImageNet compared to state-of-the-art BNN models but without extra floating-point multiplications.

机译：我们提出了BitFlow-Net，这是一种将二进制CNN模型简化为在推理时几乎没有浮点乘法的方法。近来，提出了许多二值化网络的变体，试图在用位运算代替消耗资源的浮点乘法的同时实现高精度。这些方法通常需要比例因子和BatchNorm才能获得与全精度同类产品相当的精度。但是，由于与比例因子和BatchNorm相乘，因此必须经常在浮点数据和位数据之间转换数据流。当在边缘硬件上实施时，这种转换将花费额外的资源和时间。为此，我们在先前工作的基础上进一步探索并揭示了BNN的一些基本属性，并提出了一种简化二进制网络的新方法。结果，我们的模型可以推断出其大部分数据流仍为位流。当在ASIC或FPGA上实现时，这样的网络体系结构将大大降低设计复杂性。与最新的BNN模型相比，我们的方法不会对ImageNet造成任何精度下降，但无需额外的浮点乘法。

著录项

来源
《2018 International Conference on Cloud Computing, Big Data and Blockchain》|2018年|1-6|共6页
会议地点 Fuzhou(CN)
作者
Peiqing Jiang; Lijun Wu; Zhicong Chen; Yunfeng Lai; Shuying Cheng; Peijie Lin;
展开▼
作者单位

College of Physics and Information Engineering, Fuzhou University, Fuzhou, China;

College of Physics and Information Engineering, Fuzhou University, Fuzhou, China;

College of Physics and Information Engineering, Fuzhou University, Fuzhou, China;

College of Physics and Information Engineering, Fuzhou University, Fuzhou, China;

College of Physics and Information Engineering, Fuzhou University, Fuzhou, China;

College of Physics and Information Engineering, Fuzhou University, Fuzhou, China;

展开▼
会议组织
原文格式 PDF
正文语种 eng
中图分类
关键词
computer vision; distributed processing; edge detection; fixed point arithmetic; floating point arithmetic; multiplying circuits;

机译：计算机视觉;分布式处理;边缘检测;定点算法;浮点算法;乘法电路;;

相似文献

外文文献
中文文献
专利

1. Reconfigurable Smart In-Memory Computing Platform Supporting Logic and Binarized Neural Networks for Low-Power Edge Devices [J] . Zanotti Tommaso, Puglisi Francesco Maria, Pavan Paolo Emerging and Selected Topics in Circuits and Systems, IEEE Journal on . 2020,第4期

机译：可重新配置的智能内存计算平台支持低功率边缘设备的逻辑和二金属化神经网络
2. A Survey of Convolutional Neural Networks on Edge with Reconfigurable Computing [J] . Mário P. Véstias Algorithms . 2019,第8期

机译：可重构计算的边缘卷积神经网络研究
3. Pruning deep convolutional neural networks for efficient edge computing in condition assessment of infrastructures [J] . Wu Rih-Teng, Singla Ankush, Jahanshahi Mohammad R., Computer-Aided Civil and Infrastructure Engineering . 2019,第9期

机译：修剪深度卷积神经网络以进行基础结构状态评估中的有效边缘计算
4. Fully Binarized Convolutional Neural Network for Accelerating Edge Vision Computing [C] . Peiqing Jiang, Lijun Wu, Zhicong Chen, International Conference on Cloud Computing, Big Data and Blockchain . 2018

机译：完全二金化卷积神经网络，用于加速边缘视觉计算
5. Learning from minimally labeled data with accelerated convolutional neural networks. [D] . Dundar, Aysegul. 2016

机译：通过加速卷积神经网络从最小标记的数据中学习。
6. Accelerating Inference of Convolutional Neural Networks Using In-memory Computing [O] . Martino Dazzi, Abu Sebastian, Luca Benini, 2021

机译：使用内存计算加速卷积神经网络的推断
7. Systolic-CNN: An OpenCL-defined Scalable Run-time-flexible FPGA Accelerator Architecture for Accelerating Convolutional Neural Network Inference in Cloud/Edge Computing [O] . Akshay Dua, Yixing Li, Fengbo Ren 2020

机译：Systolic-CNN：用于在云/边缘计算中加速卷积神经网络推断的OpenCL定义可伸缩的运行时柔性FPGA加速器架构

Fully Binarized Convolutional Neural Network for Accelerating Edge Vision Computing

摘要

著录项

相似文献

相关主题

期刊订阅