In the practical application scenarios of the banking and financial system, biometric technology still faces significant limitations and security risks. This proposal introduces a new multimodal identification method combined with intelligent microexpression markers based on artificial intelligence. Its goal is to enhance biometric accuracy and provide users with intelligent emotional interaction through microexpression authentication.

"Application and Analysis of Biometric Technology in Banking" is a biometric technology:

Stable Net application model: It adopts random Fourier feature reduction to improve the independence of variables (biometric features) by reducing collinearity.

Hierarchical network microexpression recognition algorithm: It can use the CASME-II dataset for pre-training VGG face, then use it as a feature processor to remove redundancies from the images. The processed images are then uploaded to the LSTM network for temporal arrangement, obtaining the user's dynamic expressions.

With the improvements in the above models, three major innovative extensions can be achieved in the field of biometric identification: solving environmental interference issues, promoting live detection upgrades, and integrating multiple domain expertise. Through microexpression recognition technology, it promotes the development of biometric identification towards deep intelligence.

Through these innovations, this project can not only be applied to banking mobile apps but also significantly enhance the security of future devices that require biometric identification.

#Industrial and Commercial Bank Cup #3rd University of Macau Financial Technology Innovation Competition #Bank of Communications Macau #University of Macau Students #Illuminate Youth with Creativity #FinTech #Biometric Technology #Bank