پیاده سازی شبکه عصبی پیچشی
ترجمه نشده

پیاده سازی شبکه عصبی پیچشی

عنوان فارسی مقاله: کاربرد سیستم شماره باقیمانده برای کاهش هزینه های سخت افزاری پیاده سازی شبکه عصبی پیچشی
عنوان انگلیسی مقاله: Application of the residue number system to reduce hardware costs of the convolutional neural network implementation
مجله/کنفرانس: ریاضیات و رایانه ها در شبیه سازی – Mathematics and Computers in Simulation
رشته های تحصیلی مرتبط: مهندسی کامپیوتر
گرایش های تحصیلی مرتبط: هوش مصنوعی، مهندسی سخت افزار
کلمات کلیدی فارسی: پردازش تصویر، شبکه عصبی پیچشی، سیستم شماره باقیمانده، نویز کمیت سازی، آرایه گیت قابل برنامه ریزی میدانی
کلمات کلیدی انگلیسی: Image processing; Convolutional neural networks; Residue number system; Quantization noise; Field-programmable gate array (FPGA
نوع نگارش مقاله: مقاله پژوهشی (Research Article)
نمایه: Scopus – Master Journals List – JCR
شناسه دیجیتال (DOI): https://doi.org/10.1016/j.matcom.2020.04.031
دانشگاه: Department of Applied Mathematics and Mathematical Modeling, North-Caucasus Federal University, Stavropol, Russia
ناشر: الزویر - Elsevier
نوع ارائه مقاله: ژورنال
نوع مقاله: ISI
سال انتشار مقاله: 2020
ایمپکت فاکتور: 1.791 در سال 2019
شاخص H_index: 67 در سال 2020
شاخص SJR: 0.526 در سال 2019
شناسه ISSN: 0378-4754
شاخص Quartile (چارک): Q2 در سال 2019
فرمت مقاله انگلیسی: PDF
تعداد صفحات مقاله انگلیسی: 12
وضعیت ترجمه: ترجمه نشده است
قیمت مقاله انگلیسی: رایگان
آیا این مقاله بیس است: خیر
آیا این مقاله مدل مفهومی دارد: ندارد
آیا این مقاله پرسشنامه دارد: ندارد
آیا این مقاله متغیر دارد: ندارد
کد محصول: E15008
رفرنس: دارای رفرنس در داخل متن و انتهای مقاله
فهرست انگلیسی مطالب

Abstract


۱٫ Introduction


۲٫ Convolutional neural networks


۳٫ Background on RNS


۴٫ Convolution in the RNS


۵٫ Software simulation


۶٫ Hardware implementation


۷٫ Conclusion


Acknowledgment


References

نمونه متن انگلیسی مقاله

Abstract


Convolutional neural networks are a promising tool for solving the problem of pattern recognition. Most well-known convolutional neural networks implementations require a significant amount of memory to store weights in the process of learning and working. We propose a convolutional neural network architecture in which the neural network is divided into hardware and software parts to increase performance and reduce the cost of implementation resources. We also propose to use the residue number system (RNS) in the hardware part to implement the convolutional layer of the neural network. Software simulations using Matlab 2018b showed that convolutional neural network with a minimum number of layers can be quickly and successfully trained. The hardware implementation of the convolution layer shows that the use of RNS allows to reduce the hardware costs on 7.86%–۳۷٫۷۸% compared to the two’s complement implementation. The use of the proposed heterogeneous implementation reduces the average time of image recognition by 41.17%.


Introduction


The modern development of science and technology implies the widespread introduction of data mining methods. The range of tasks requiring the use of artificial intelligence methods in the field of image processing is constantly expanding: personal identification [5], scene recognition [6], information processing from external sensors in unmanned land and aircraft vehicles [28], medical diagnostics [23], and so on. The use of intelligent methods based on artificial neural networks is a promising tool for solving the problem of image recognition [35]. The idea of using artificial neural networks for processing visual information was proposed in [15] to solve the problem of automating the recognition of handwritten numbers. The architecture proposed in this paper was called the Convolutional Neural Network (CNN). The combination of convolutional layers, realizing the extraction of visual signs, and a multilayer perceptron, realizing the recognition operation according to the convolution results, is its main feature. The evolution of this scientific idea and the development of computer technology have led to the fact that at present the theory of CNN and its practical application methods are developing along the path of an extensive increase in the number of layers of CNN, which leads to the high computational complexity of the implementation of such systems. For example, the network architecture [14], which showed the best ImageNet image recognition result in 2010, contains about 650 thousand neurons, 60 million tunable parameters and requires 27 GB of disk space for training. The paper [30] presents the development of Google, which showed the best image recognition result of ImageNet base in 2014. This CNN performs more than one and a half billion computational operations for image recognition, which motivated Google to develop a special tensor processing unit to optimize the operation of this CNN [13]. Summarizing, modern CNN architectures are very resource intensive, which limits the possibilities for their wide practical application. A unified approach to reducing resource costs in the implementation of CNN in practice is not currently available.

  • اشتراک گذاری در

دیدگاه خود را بنویسید:

تاکنون دیدگاهی برای این نوشته ارسال نشده است

پیاده سازی شبکه عصبی پیچشی
نوشته های مرتبط
مقالات جدید
لوگوی رسانه های برخط

logo-samandehi

پیوندها