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

A Configurable Neural Network Reasoning and Online Learning Fusion Computing Circuit

A technology of neural network and computing unit, applied in the field of neural network, can solve problems such as missing and imperfect online learning functions, and achieve the effect of improving accuracy, improving efficiency, and improving utilization rate

Active Publication Date: 2020-12-29
UNIV OF ELECTRONICS SCI & TECH OF CHINA
View PDF10 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In view of the imperfect or even lack of online learning functions in the traditional technical solutions, the present invention proposes a configurable neural network reasoning and online learning fusion computing circuit

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 Configurable Neural Network Reasoning and Online Learning Fusion Computing Circuit
  • A Configurable Neural Network Reasoning and Online Learning Fusion Computing Circuit
  • A Configurable Neural Network Reasoning and Online Learning Fusion Computing Circuit

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0035] In order to make the purpose, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with the implementation methods and accompanying drawings.

[0036] figure 1 For this embodiment, a schematic diagram of a neural network reasoning and online learning fusion computing circuit module can be configured, such as figure 1 As shown, the circuit includes a main control unit, a configurable control unit, a cache control unit, a compression / decompression unit, an interaction unit, a cache unit, and a neural network fusion calculation unit. Among them, the neural network fusion computing unit includes convolution multiplication and addition unit, forward reasoning key computing unit and backpropagation key computing unit. The key calculation unit of forward inference and the key calculation unit of backpropagation respectively form the inference calculation path and the backpropagation ca...

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 configurable neural network reasoning and online learning fusion computing circuit, which belongs to the technical field of neural network. The circuit includes a main control unit, a configurable control unit, a cache control unit, a compression / decompression unit, an interaction unit, a cache unit, and a neural network fusion calculation unit. The circuit of the present invention realizes low-cost configurable reasoning and online learning fusion calculation by configuring the calculation path of the neural network, the number of layers of online learning, and the label generation method, ensuring efficient reasoning and learning of the neural network; through hardware-based dynamic complex The calculation engine is used to integrate reasoning and online learning calculations, which improves the utilization of hardware.

Description

technical field [0001] The invention belongs to the field of neural networks and relates to a configurable neural network reasoning and online learning fusion computing circuit. Background technique [0002] As an important branch of artificial intelligence, Deep Neural Network (DNN) has received great attention because it can assist or replace humans in the fields of face recognition, speech recognition, natural language processing, and medical aided diagnosis. Among them, many researchers have conducted extensive research on the reasoning and learning (training) of neural networks. The reasoning of the neural network can be regarded as further analysis and inference of the input data, and accurate inference is inseparable from the effective learning of the network parameters by the neural network. Generally, the learning of neural networks can be divided into offline learning and online learning. Offline learning only uses centralized offline data for learning, and the p...

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 Patents(China)
IPC IPC(8): G06N3/063G06N3/08G06N5/04G06F7/544
CPCG06F7/5443G06N3/063G06N3/084G06N5/046
Inventor 周军甘家彦弓育川
Owner UNIV OF ELECTRONICS SCI & TECH OF CHINA