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

Face image illumination migration method based on convolutional neural network

A convolutional neural network and light transfer technology, applied in neural learning methods, biological neural network models, neural architectures, etc., can solve problems such as lighting that does not fully meet practical application requirements, and that it does not deal with face image neck and background lighting.

Active Publication Date: 2020-05-01
韶鼎人工智能科技有限公司
View PDF9 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

At present, the lighting migration methods of face images only change the lighting of the face of the face image, but do not deal with the lighting of the neck and background of the face image.
In actual application scenarios, the face part and the non-face part are inseparable, and only changing the illumination of the face part does not fully meet the actual application requirements.

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
  • Face image illumination migration method based on convolutional neural network
  • Face image illumination migration method based on convolutional neural network
  • Face image illumination migration method based on convolutional neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0039] In order to better understand the present invention, some basic concepts are firstly explained.

[0040] Convolutional Neural Networks: Convolutional Nerual Networks, CNN is a type of neural network that includes convolutional calculations and is often used in deep learning.

[0041] Convolution layer: extract image features through convolution calculation;

[0042] Pooling layer: Compress the input feature image to simplify the computational complexity of the network;

[0043] Fully connected layer: connect all the features and send the output value to the softmax layer;

[0044] Migration learning: Migrate the weights in a network model that has been trained to complete a certain classification task to a new network model trained for another target classification, instead of training from the initial state.

[0045] Illumination matching: Illumination matching is given the face illumination input image, looking for an image with the same illumination effect as the i...

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 face image illumination migration method based on a convolutional neural network. Illumination migration of a face image is realized by using the convolutional neural networkCNN. The method is mainly realized by two parts: illumination model training and illumination classification, illumination matching and illumination migration realized on the basis of reference imagestyle migration. Firstly, illumination classification is completed on a Yale Face face data set and a PIE face data set by combining convolutional neural networks VGG19 and VGG16, and a model capableof classifying face image illumination is obtained; then, the model is utilized to realize illumination matching of a single face image, and an image similar to a given face image in illumination canbe obtained from a face image illumination data set; and finally, through an illumination classification model, extracting and processing related illumination features of a given reference face illumination image so as to facilitate migration to an input face image, thereby realizing overall migration of illumination of a single face image.

Description

technical field [0001] The invention is a convolutional neural network-based illumination migration method for a face image, which belongs to the field of computer vision. Background technique [0002] The lighting effect of the image is a research hotspot in many research directions in the field of computer vision, and it is also very important for the portrait image itself. The effect of light and shadow has a wide range of application requirements in modern digital film and television production, portrait photography beautification, advertising and other art design. One of the difficult problems in film and television production is how to capture the best lens performance of actors and at the same time, the lighting effect in the scene is the most ideal effect. To solve this problem from the aspect of improving the shooting process, it is faced with expensive costs, and to solve this problem from the aspect of post-production, it is faced with complex technical processin...

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): G06T3/00G06K9/62G06N3/04G06N3/08
CPCG06N3/08G06N3/045G06F18/241G06T3/04
Inventor 金鑫李忠兰肖超恩
Owner 韶鼎人工智能科技有限公司