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

Feature extraction network training method, image processing method, device and equipment thereof

A feature extraction and network training technology, applied in the field of computer vision, can solve problems such as performance degradation

Active Publication Date: 2021-09-14
SHANGHAI SENSETIME LINGANG INTELLIGENT TECH CO LTD
View PDF8 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] However, in the real world, this basic assumption is likely to be broken, resulting in trained network models such as machine learning / deep learning. When using test data from different domains from the training data, machine learning / deep learning The learning model will show different degrees of performance degradation

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
  • Feature extraction network training method, image processing method, device and equipment thereof
  • Feature extraction network training method, image processing method, device and equipment thereof
  • Feature extraction network training method, image processing method, device and equipment thereof

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0046] The exemplary embodiment will be described in detail herein, and examples thereof are shown in the drawings. The following description is related to the drawings, unless otherwise indicated, the same numbers in the drawings represent the same or similar elements. The embodiments described in the exemplary embodiments described below do not represent all embodiments consistent with the present disclosure. Instead, they are only examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.

[0047] The technical solutions of the present invention and how the technical solutions of the present invention will be described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and will not be described in some embodiments for the same or similar concepts or processes. Embodiments of the present invention will be described below with reference to the accompanying d...

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

Embodiments of the present invention provide a feature extraction network training method, image processing method, device and equipment thereof. The training method includes: acquiring a source domain image set and a target domain image set, and the target domain image set includes less than The number of source domain images included in the source domain image set; the source domain image and the target domain image are input to the feature extraction network for feature extraction respectively, and the first feature information of the source domain image and the second feature information of the target domain image are obtained; according to the first The first feature information and the second feature information respectively determine the domain categories corresponding to the source domain image and the target domain image; at least based on the confrontation between the domain category determination results of the source domain image and the target domain image and the real domain category, adjust the feature extraction The network parameters of the network.

Description

Technical field [0001] Embodiments of the present invention relates to computer vision, and more particularly, to a method of network training feature extraction, image processing method, apparatus and equipment. Background technique [0002] Currently used by the machine learning / deep learning models are based on a common basis for the assumption that the training data (training data) distribution profile (distribution) and the test data (testing data) are the same, or at least similar. In this way, machine learning / learning model in depth on the training data to learn the parameters, it can be applied differently to training data test data. [0003] However, in the real world, this basic hypothesis is likely to be broken, resulting in machine learning / depth learning network model training obtained, and training in the use of data from the test data from different fields of machine learning / depth learning model will have different degrees of performance degradation. In o...

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): G06K9/00G06K9/46G06N3/04G06N3/08
CPCG06N3/08G06V20/00G06V10/462G06V2201/08G06V2201/07G06N3/045
Inventor 苏鹏王坤曾星宇
Owner SHANGHAI SENSETIME LINGANG INTELLIGENT TECH CO LTD