Unlock instant, AI-driven research and patent intelligence for your innovation.

Convolutional neural network acceleration device and method

A conversion device and one-line technology, applied in the field of convolutional neural network, can solve the problems of reducing the throughput rate of convolutional neural network, achieve the effect of reducing computing power requirements, reducing memory bandwidth requirements, and reducing burden

Active Publication Date: 2018-08-10
SHANGHAI THINK FORCE ELECTRONICS TECH CO LTD
View PDF6 Cites 11 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

This will lead to increased bandwidth requirements for input data, or lower throughput of convolutional neural networks at the same bandwidth

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Convolutional neural network acceleration device and method
  • Convolutional neural network acceleration device and method
  • Convolutional neural network acceleration device and method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0044] In the following description, the present invention is described with reference to various examples. One skilled in the art will recognize, however, that the various embodiments may be practiced without one or more of the specific details, or with other alternative and / or additional methods, materials, or components. In other instances, well-known structures, materials, or operations are not shown or described in detail so as not to obscure aspects of the various embodiments of the invention. Similarly, for purposes of explanation, specific quantities, materials and configurations are set forth in order to provide a thorough understanding of embodiments of the invention. However, the invention may be practiced without these specific details. Furthermore, it should be understood that the various embodiments shown in the drawings are illustrative representations and are not necessarily drawn to scale.

[0045] In this specification, reference to "one embodiment" or "the...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The embodiment of the invention provides a convolutional neural network acceleration device and method. The convolutional neural network acceleration device comprises a scheduling unit, a data movingunit, a row data caching unit and a row data extension unit, wherein the scheduling unit is used for producing control commands according to sizes / dimensions of input data and sizes / step lengths of awave filter; the data moving unit is used for actively reading original input data from a system storage space according to the commands of the scheduling unit; the row data caching unit is used for storing the read-in original input data; the row data extension unit is used for reading a row of original data from the row data caching unit every time and then extending row data into the row of data in different filtering windows according to the sizes of the filter windows.

Description

technical field [0001] The present invention relates to the field of convolutional neural networks, in particular to a convolutional neural network acceleration device and method. Background technique [0002] Convolutional Neural Network (CNN) is a feed-forward neural network. Compared with the traditional BP neural network, it has the advantages of high recognition efficiency and good rotation and scaling invariance. It has been used in digital image processing and face recognition. It has been widely used in various fields. [0003] Traditional convolutional neural networks generally consist of multiple alternating convolutional layers, pooling layers, and finally fully connected layers. Convolutional neural networks can pass the network loss to all layers of the network through the backpropagation method. The parameter update learning process is realized by the stochastic gradient descent algorithm. The biggest difference between the convolutional neural network and t...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06F15/78G06N3/063
CPCG06F15/781G06F15/7825G06N3/063G06N3/045G06F9/4881G06N3/04G06F13/4018
Inventor 刘明润陈亮李晓鹏
Owner SHANGHAI THINK FORCE ELECTRONICS TECH CO LTD