Network structure setting method for secondary face detection
A network structure and face detection technology, which is applied in the field of face image recognition, can solve problems such as false detection, failure to use normally, and decline in the recognition rate of face recognition, so as to improve efficiency, increase recognition rate, and reduce the amount of calculation.
Active Publication Date: 2020-11-27
北京君正集成电路股份有限公司
View PDF4 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Its disadvantages include: there are certain false detections in the front-end face detection, the detected face frame and the judgment standard of whether it is a face do not match the requirements of face recognition, resulting in a decline in the recognition rate of face recognition or even failure to use normally
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
[0034] In the field of face recognition, some current terms in related technical fields include:
[0035] 1. Deep learning: The concept of deep learning originated from the research of artificial neural networks. A multi-layer perceptron with multiple hidden layers is a deep learning structure. Deep learning combines low-level features to form more abstract high-level representation attribute categories or features to discover distributed feature representations of data.
[0036] 2. Face detection: The process of using a face detector to detect whether there is a face in a video or a picture is called face detection.
[0037] 3. Convolution kernel: The convolution kernel is a matrix used for image processing, and a parameter for calculation with the original image. The convolution kernel is usually composed of a column matrix (for example, a 3*3 matrix), and each square in the area has a weight value. The matrix shape is generally 1×1, 3×3, 5×5, 7×7, 1×3, 3×1, 2×2, 1×5, 5×1...
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 provides a network structure setting method for secondary face detection, and the method comprises the steps: a first layer inputs a large-size picture, a feature map with the depth of 4is adopted, the size of a convolution kernel is 3*3, the step length is 2, and two ends of a convolution calculation image is non-aligned; in a second layer, a picture with an appropriate size is input, a feature map with the output depth of 8 is adopted, the convolution kernel size is 3*3, the step length is 2, and the two ends of convolution calculation are not aligned; a third layer inputs a picture with a small size, a feature map with the output depth of 16 is adopted, the size of a convolution kernel is 3*3, the step length is 2, two ends of convolution calculation are not aligned, andan output feature map with the size of 6*6*16 is obtained, the input data of a fourth layer is output of the third layer, 32 feature maps are output, the size of a convolution kernel is 1*1, the steplength is 1, the two ends of convolution calculation are aligned, and an output feature map with the size of 6*6* 32 is obtained, the input data of a fifth layer is the output of the fourth layer, 64feature maps are output, the size of a convolution kernel is 3*3, the step length is 2, two ends of convolution calculation are aligned, and an output feature map with the size of 3*3*64 is obtained;and a sixth layer is fully connected to a feature map of 192.
Description
technical field [0001] The invention relates to the technical field of face image recognition, in particular to a network structure setting method for secondary face detection. Background technique [0002] With the continuous development of science and technology, especially the development of computer vision technology, face recognition technology is widely used in various fields such as information security and electronic authentication, and the image feature extraction method has good recognition performance. Face recognition refers to the technology of identifying one or more faces from static or dynamic scenes by using image processing and / or pattern recognition technology based on the known face sample library. But the current face recognition technology includes 1. Face detection of traditional machine learning. 2. Face detection based on deep learning. 3. It is used for front-end face detection. Its disadvantages include: there are certain false detections in the...
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): G06K9/00
CPCG06V40/161G06V40/168Y02D10/00
Inventor 于晓静田凤彬
Owner 北京君正集成电路股份有限公司