Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

A three-dimensional convolution neural network implementation method based on memristor

A neural network and three-dimensional convolution technology, applied in the direction of biological neural network models, neural architectures, etc., can solve problems such as limited memristor characteristics, achieve hardware noise resistance, reduce precision and preparation difficulty, and satisfy video processing accuracy Effect

Active Publication Date: 2018-12-25
重庆因普乐科技有限公司
View PDF7 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] However, most of the neural networks implemented based on memristors are the most basic fully connected layers or simple convolutional layers, which limits the use of memristor characteristics, and the existing research is limited to the recognition of simple signal and image models.

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
  • A three-dimensional convolution neural network implementation method based on memristor
  • A three-dimensional convolution neural network implementation method based on memristor
  • A three-dimensional convolution neural network implementation method based on memristor

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0032] The specific embodiment and working principle of the present invention will be described in further detail below in conjunction with the accompanying drawings.

[0033] Such as figure 1 As shown, a memristor-based three-dimensional convolutional neural network implementation method, the specific steps are as follows:

[0034]Step 1: Based on the basic memristor array, prepare a network consisting of the first convolution + pooling layer, the second convolution + pooling layer and the fully connected layer. figure 2 The three-dimensional convolutional neural network to be trained is shown, and the original video information is input to obtain the output value of the fully connected layer;

[0035] The specific preparation process of the three-dimensional convolutional neural network to be trained is as follows:

[0036] Step 1.1: Prepare the first convolution + pooling layer memristor array using the basic memristor array, and input the original video information to o...

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 method for realizing a three-dimensional convolution neural network based on a memristor, which comprises the following steps: preparing a three-dimensional convolution neural network to be trained composed of a first convolution+pooling layer, a second convolution+pooling layer and a full connection layer, and inputting original video information to obtain an output value of the full connection layer; according to the deviation between the output value of all connection layers of the three-dimensional convolution neural network to be trained and the standard information, the back propagation function being used to train the three-dimensional convolution neural network to be trained; when the set training times are reached, whether the training precision reaches the standard or not being judged; if the training precision does not reach the standard, the training being carried out again until the precision reaches the standard; and obtaining the required three-dimensional convolution neural network and the like. The remarkable effects are: the influence on precision, difficulty of preparation and time is reduced in hardware; it is easy to build and implement hardware, at the same time, the difficulty of neural network based on memristor is raised to video level for the first time.

Description

technical field [0001] The invention relates to the technical fields of artificial intelligence and material science, in particular to a method for realizing a three-dimensional convolutional neural network based on a memristor. Background technique [0002] Memristor is a dimension with resistance, as the fourth basic element following the resistor representing the relationship between voltage and current, the capacitor representing the relationship between charge and voltage, and the inductor representing the relationship between magnetic flux and current. A circuit device that represents the relationship between magnetic flux and charge (dρ = Mdq). The resistance value of the memristor will change with the amount of current flowing through the memristor. Even if the current stops, its resistance value will still stop at the value before the current stops, unless the current flows through the memristor again. For memristors, if the current flows in the forward direction, ...

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): G06N3/04
CPCG06N3/045
Inventor 李正浩刘佳琪唐永亮李靖禾
Owner 重庆因普乐科技有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products