Face recognition method and device based on sequential neural network model
A neural network model and face recognition technology, which are applied in the field of face recognition methods and devices based on sequential neural network models, can solve the problem that face recognition technology cannot meet practical requirements, changes in lighting, changes in user posture and expression, and changes in age and body shape. and other problems to achieve the effect of improving generalization ability, saving computing time, and improving accuracy
Active Publication Date: 2019-08-23
INST OF AUTOMATION CHINESE ACAD OF SCI
View PDF2 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
At present, face recognition technology still cannot meet the practical requirements in an outdoor uncontrolled environment. The main difficulties lie in changes in lighting, changes in user posture and expression, changes in age and body shape, and occlusion.
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 example
[0061] In order to describe the specific implementation of the present invention and verify the effectiveness of the present invention, we apply the method proposed by the present invention to a public face database——LFW face database. The database includes 5749 individuals with a total of 13233 images.
[0062] In our example, we adopt the standard testing protocol on the LFW dataset to demonstrate the effectiveness of the present invention. The standard test protocol for the LFW dataset consists of 6000 pairs of face images, including 3000 pairs of face images of the same person and 3000 pairs of face images of different people.
[0063] Specific steps are as follows:
[0064] Training process:
[0065] Step S3-1, collecting a large number of face images as training data, and designing a neural network model. In particular, the neural network model we use contains 4 convolutional layers and 4 pooling layers. After each pooling layer, the output is divided into two groups,...
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 face recognition method and device based on a sequential neural network model. The method includes: preprocessing the input face image, correcting the angle and expression of the face image; extracting the features of the corrected face image / video by using a neural network including sequencing operations; Calculate the similarity between image pairs to know the identity of a specific object in the input face image. In the face recognition problem, the face recognition model based on the neural network has many parameters and the problem of high calculation cost, and proposes a sequenced neural network structure, which effectively reduces network parameters and saves calculation time through the sequenced representation between different features ; and for the problem of less training data, a training method based on contrastive loss and triplet loss is proposed.
Description
technical field [0001] The invention relates to technical fields such as artificial intelligence, pattern recognition, and digital image processing, and in particular to a face recognition method and device based on a sequential neural network model. Background technique [0002] As a kind of biometric identification technology, face recognition has good development and application prospects due to its non-contact and convenient collection characteristics. Face recognition technology has played a very important role in many application scenarios, such as airport security check, border inspection and customs clearance. In recent years, with the rapid development of Internet finance, face recognition technology has shown great application advantages in mobile payment. The purpose of face recognition is to know the user's identity based on the acquired user's face image or video. At present, face recognition technology still cannot meet the practical requirements in an outdoo...
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 Patents(China)
IPC IPC(8): G06K9/00
CPCG06V40/165G06V40/168
Inventor 孙哲南赫然谭铁牛宋凌霄曹冬侯广琦
Owner INST OF AUTOMATION CHINESE ACAD OF SCI