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

Classification system and method for dynamically updating convolutional neural network based on ZYNQ

A convolutional neural network, dynamic update technology, applied in the field of communication, to achieve the effect of small equipment and low power consumption

Active Publication Date: 2020-07-17
UNIV OF ELECTRONIC SCI & TECH OF CHINA
View PDF6 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] The purpose of the present invention is to solve the problem of updating and classifying convolutional neural networks based on ZYNQ, and proposes a classification system and method for dynamically updating convolutional neural networks based on ZYNQ

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
  • Classification system and method for dynamically updating convolutional neural network based on ZYNQ
  • Classification system and method for dynamically updating convolutional neural network based on ZYNQ
  • Classification system and method for dynamically updating convolutional neural network based on ZYNQ

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0058] Embodiments of the present invention will be further described below in conjunction with the accompanying drawings.

[0059] Such as figure 1 As shown, the present invention provides a kind of classification system based on ZYNQ dynamic update convolution neural network, including ZYNQ chip 1, Gigabit Ethernet port 2, FEP interface 3, DDR memory 4, server 5 and FEP interface camera 6; Said ZYNQ chip 1 is connected with Gigabit Ethernet port 2, FEP interface 3 and DDR memory 4 respectively; said Gigabit Ethernet port 2 is connected with server 5; said FEP interface 3 is connected with FEP interface camera 6.

[0060] In the embodiment of the present invention, such as figure 2 As shown, ZYNQ chip 1 includes FPGA7, ARM dual-core processor 8, first bus 9, second bus 10 and DDR controller 11; FPGA7 includes register 12, SRAM cache 13, DMA access module 14, data receiving module 15, pool Transformation circuit 16 and convolution multiplication and addition circuit 17 are ...

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 invention discloses a classification system and method for dynamically updating a convolutional neural network based on ZYNQ. The classification system comprises a ZYNQ chip, a gigabit Ethernet port, an FEP interface, a DDR memory, a server and an FEP interface camera. The ZYNQ chip is in communication connection with the gigabit Ethernet port, the FEP interface and the DDR memory. The gigabitEthernet port is in communication connection with the server; and the FEP interface is in communication connection with the FEP interface camera. The method supports the dynamic updating of the typesand parameters of the convolutional neural network, supports the online smooth updating, and does not affect the simultaneous execution of an updating task and a classification task. And meanwhile, the device can be applied to complex environments such as equipment with requirements on power consumption or fields. The system can also support a plurality of ZYNQ devices at the same time, and can update a plurality of ZYNQs concurrently.

Description

technical field [0001] The invention belongs to the technical field of communication, and in particular relates to a classification system and method for dynamically updating a convolutional neural network based on ZYNQ. Background technique [0002] With the advancement of technology, the classification system based on ZYNQ's convolutional neural network is getting faster and faster. However, the traditional ZYNQ-based convolutional neural network classification system is only for a specific convolutional neural network algorithm. Difficult to support dynamic updates, difficult to cope with complex and changeable targets. At the same time, the power consumption of the convolutional neural network classification system based on CPU and GPU is too large, and it is difficult to apply it to harsh environments such as the wild in the form of equipment. Based on the above situation, the present invention proposes a classification system and method based on ZYNQ to dynamically u...

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
IPC IPC(8): G06F15/177G06F13/32G06F9/54G06K9/62G06N3/04G06K9/00
CPCG06F15/177G06F13/32G06F9/544G06V10/955G06N3/045G06F18/24Y02D10/00
Inventor 刘禹兵潘晔邵怀宗胡全林静然利强
Owner UNIV OF ELECTRONIC SCI & TECH OF CHINA