A method of identity recognition based on dual-channel convolutional neural network

Abstract

The invention discloses an identity recognition method based on a dual-channel convolutional neural network, which includes training the neural network and identity recognition steps, using two time-synchronized images of a face image and a body posture image for comprehensive training and recognition, avoiding a single factor Deception has strong anti-interference ability and high recognition accuracy; the method of the present invention weights the feature data of two channels through a fully connected layer, and obtains image features through multiple convolutional layers and pooling layers Finally, the classification probability is obtained by the classifier, and the maximum probability is extracted to compare with the set threshold to determine the recognition result. After multiple convolutions to extract feature maps, nonlinear excitation and pooling dimensionality reduction processing, the dual-channel convolutional neural network control data is more flexible, and the abstraction ability and learning ability are stronger, thus having a better recognition effect.

Description

technical field [0001] The invention relates to the technical field of artificial intelligence, in particular to an identity recognition method based on a dual-channel convolutional neural network. Background technique [0002] With the continuous development of artificial intelligence technology, face recognition technology is becoming more and more mature. Face recognition has been widely used in government, military, bank, Social welfare, e-commerce, security and defense and other fields. However, because the shape of the face is very unstable, people can produce many expressions through changes in the face, and the visual images of the face are also very different from different viewing angles, so the current face recognition technology also has many limitations. In addition, face recognition is also easily affected by various external environments and interference conditions. For example, the collection of portraits is sensitive to the surrounding light environment, an...

AI Technical Summary

Problems solved by technology

However, because the shape of the face is very unstable, people can produce many expressions through changes in the face, and the visual images of the face are also very different from different viewing angles, so the current face recognition technology also has many limitations.

In addition, face recognition is also easily affected by various external environments and interference conditions. For example, the collection of portraits is sensitive to the surrounding light environment, and the difference in light will greatly affect the accuracy of recognition. Factors such as various ornaments and other occluders and face aging also have a greater impact on the recognition rate

Images