Portrait segmentation model training method, storage medium and terminal equipment
A segmentation model and training method technology, applied in image analysis, image enhancement, image data processing and other directions, can solve the problems of uneven edge fitting, poor portrait segmentation effect, etc., to avoid unsmooth fitting effect and improve accuracy. Effect
Pending Publication Date: 2021-08-13
WUHAN TCL CORP RES CO LTD
View PDF3 Cites 1 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
At present, the deep convolutional neural network model commonly used in portrait segmentation technology has unsmooth edge fitting, which makes the portrait segmentation effect poor.
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 moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0041] The present invention provides a training method, storage medium and terminal equipment for a portrait segmentation model. In order to make the purpose, technical solution and effect of the present invention clearer and clearer, the present invention will be further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described here are only used to explain the present invention, not to limit the present invention.
[0042] Those skilled in the art will understand that unless otherwise stated, the singular forms "a", "an", "said" and "the" used herein may also include plural forms. It should be further understood that the word "comprising" used in the description of the present invention refers to the presence of said features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, Integers, steps, operations...
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
Login to View More Abstract
The invention discloses a portrait segmentation model training method, a storage medium and terminal equipment, and the method comprises the steps: extracting a predicted portrait image and a predicted edge image of a training image through a to-be-trained portrait segmentation model; determining a loss function corresponding to the training image based on the predicted portrait image, a real portrait image corresponding to the training image, the predicted edge image and a real edge image corresponding to the training image; and training the to-be-trained portrait segmentation model based on the loss function. The predicted portrait image and the predicted edge image of the training image are extracted through the to-be-trained portrait segmentation model, and the portrait segmentation model is trained through the predicted portrait image and the predicted edge image, so that the problems of unsmooth fitting effect and wrong segmentation at the boundary can be avoided, therefore, the precision of the portrait segmentation model obtained by training is improved.
Description
technical field [0001] The present invention relates to the technical field of portrait segmentation, in particular to a training method, storage medium and terminal equipment for a portrait segmentation model. Background technique [0002] In the process of image processing, it may be necessary to blur the image background, and in the process of processing the image background, it is necessary to separate the human face from the image background through portrait segmentation technology. With the continuous development of deep learning technology, the deep convolutional neural network model is also used in the portrait segmentation technology to realize the image segmentation of the portrait area. At present, the deep convolutional neural network model commonly used in portrait segmentation technology has unsmooth edge fitting, which makes the portrait segmentation effect poor. [0003] Thereby prior art still needs to improve and improve. Contents of the invention [00...
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
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): G06T7/11G06T5/00G06T7/13G06T7/194
CPCG06T7/11G06T7/13G06T7/194G06T2207/20081G06T5/70
Inventor 孟俊彪
Owner WUHAN TCL CORP RES CO LTD