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

Data compression device and data compression method for deep neural network

A deep neural network and data compression technology, which is applied in the field of data compression devices and data compression, can solve problems affecting hardware costs, memory usage, and bandwidth consumption.

Pending Publication Date: 2022-04-01
BEIJING JINGSHI INTELLIGENT TECH CO LTD
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] The scale of the deep neural network will affect the hardware cost, such as memory usage and bandwidth consumption will increase with the scale of the deep neural network

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
  • Data compression device and data compression method for deep neural network
  • Data compression device and data compression method for deep neural network
  • Data compression device and data compression method for deep neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0019] In order to make the above-mentioned features and effects of the present invention more clear and understandable, the following specific examples are given together with the accompanying drawings for detailed description as follows.

[0020] The present invention does not show all possible embodiments, and other implementations not presented in the present invention may also be applicable. Furthermore, the size ratios in the drawings are not drawn to the same proportions as the actual products. Therefore, the specification and illustrations are only used to describe the embodiments, not to limit the protection scope of the present invention. In addition, the descriptions in the embodiments, such as detailed structures, process steps and material applications, etc., are for illustration purposes only, and are not intended to limit the protection scope of the present invention. The details of the steps and structure of the embodiment can be changed and modified according...

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 data compression method of a deep neural network. The data compression method includes the following steps. Remapping the plurality of original data according to at least one offset value and a positive and negative value to obtain a plurality of mapped data; a distribution center of the mapped data is aligned to 0, and the mapped data are all non-negative integers. A plurality of data blocks of the mapped data are encoded using at least two encoding modes to generate an encoded data.

Description

technical field [0001] The present invention relates to a data compression device and a data compression method, and in particular to a data compression device and a data compression method of a deep neural network. Background technique [0002] Deep Neural Networks (DNN) can be applied in different fields, such as image recognition and speech recognition, to solve various problems. In the application of deep neural network, high-performance deep neural network hardware accelerator and related hardware are required to achieve its performance. [0003] The scale of the deep neural network will affect the hardware cost, such as memory usage and bandwidth consumption will increase with the scale of the deep neural network. Therefore, how to compress the data of the deep neural network to reduce memory usage and bandwidth consumption is one of the focuses of the industry's efforts. Contents of the invention [0004] The present invention relates to a data compression device ...

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): H03M7/30G06N3/08
CPCG06N3/063G06N3/04
Inventor 邓书纬
Owner BEIJING JINGSHI INTELLIGENT TECH CO LTD