Method for training appearance evaluation model and pushing beauty parameters of appearance evaluation model and related device

A technology for evaluating models and training methods, applied in the field of live broadcast, which can solve problems such as difficult for anchor users to determine beauty parameters

Pending Publication Date: 2021-01-05
GUANGZHOU HUYA TECH CO LTD
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Although it is possible to realize that the displayed face image changes according to the change of the beautification parameters, it is difficult for the anchor user to determine the appropriate beautification parameters

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
  • Method for training appearance evaluation model and pushing beauty parameters of appearance evaluation model and related device
  • Method for training appearance evaluation model and pushing beauty parameters of appearance evaluation model and related device
  • Method for training appearance evaluation model and pushing beauty parameters of appearance evaluation model and related device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0104] Figure 1A It is a flow chart of a method for training an appearance evaluation model provided in Embodiment 1 of the present invention, Figure 1B It is a schematic structural diagram of an appearance evaluation model provided in Embodiment 1 of the present invention, Figure 1C It is a flow chart of another method for training an appearance evaluation model provided by Embodiment 1 of the present invention. This embodiment is applicable to the situation of image processing for human face and appearance. Specifically, the appearance evaluation model may be used to determine the appearance value of the human face and appearance in the image data. Further, this embodiment will describe the training method of the appearance evaluation model, which can be executed by the training device of the appearance evaluation model.

[0105] In this embodiment, during the training process, the appearance evaluation model includes a feature extractor, a regressor and a classifier. I...

Embodiment 2

[0143] figure 2 It is a flow chart of the training method of the appearance evaluation model provided by the second embodiment of the present invention. On the basis of the foregoing embodiments, the description of attribute information is added. Specifically, the training method includes the following steps:

[0144] S210. Determine sample image data.

[0145] In this embodiment, the sample image data is associated with attribute information. Exemplarily, the attribute information may represent attributes of faces in the sample image data. If the attribute information can be regional attribute information, it can be used to indicate that the face in the sample image data is a face of a person belonging to a certain region. For another example, the attribute information may be gender information, which may be used to indicate that the faces in the sample image data belong to boys or girls. In this embodiment, the sample image data associated with the same attribute inform...

Embodiment 3

[0158] image 3 It is a flow chart of a method for pushing beauty parameters provided by Embodiment 3 of the present invention. This embodiment is applicable to the situation of image processing for human face and appearance. Specifically, it is used to push the beautification parameter to the host client. The beautification parameter is a parameter used in image processing and can be used to adjust the The facial appearance of the person is determined. Exemplarily, the beautification parameters can be used for adjustments including: face shaping, face beautification, and face swapping. Among them, face shaping can be face-lifting, nose height adjustment, eye size adjustment, etc.; face beautification can be whitening, red eye removal, makeup, etc.; face changing means replacing the face image of the anchor user with another face image. Exemplarily, corresponding to face thinning, the beauty parameter is the distance between the pixel point where the face contour is located ...

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 embodiment of the invention discloses a method for training a appearance evaluation model, a method for pushing beauty parameters of the appearance evaluation model and a related device. The training method comprises the steps of determining sample image data; inputting the sample image data into a preset feature extractor for processing so as to output face features; inputting the face features into a preset classifier for processing so as to output a first appearance level; inputting the face features into a preset regression device for processing so as to output a first appearance value; training a feature extractor, a classifier and a regression device according to the first appearance level and the first appearance value; when the training is finished, outputting the feature extractor and the regression device as a face appearance evaluation model, so that the problem that proper beautifying parameters are difficult to determine in image processing for adjusting the face appearance is solved, the evaluation of the face appearance value of the face in the image data is realized, and further, an evaluation criterion can be provided for adjusting the face in the image data bydetermining the appearance value.

Description

technical field [0001] The embodiments of the present invention relate to live broadcast technology, and in particular to methods and related devices for training an appearance evaluation model and pushing beauty parameters. Background technique [0002] On the live broadcast platform, the anchor user can use the face processing tool provided by the live broadcast platform to perform image processing on the face image in the live video stream corresponding to the program during the live broadcast of the program, so as to realize the adjustment of the live broadcast of the anchor user. Appearance in video stream. Specifically, the functions realized by the adjustment may include: face shaping, face beautification, face changing, and the like. Among them, face shaping can be face-lifting, nose height adjustment, eye size adjustment, etc.; face beautification can be whitening, red eye removal, makeup, etc.; face changing means replacing the face image of the anchor user with a...

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): G06K9/00
CPCG06V40/172G06V40/168
Inventor 任高生王文斓
Owner GUANGZHOU HUYA TECH CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap