Enhancing societal security: a multimodal deep learning approach for a public person identification and tracking...

Area of Science:

• Computer Science
• Artificial Intelligence
• Biometrics

Background:

• Public spaces face significant security threats.
• Emerging technologies offer potential solutions for societal security.
• Intelligent person identification is crucial for monitoring and security.

Purpose of the Study:

• To develop and evaluate an intelligent person identification system for public spaces.
• To enhance security through a multimodal recognition approach.
• To investigate the integration of deep learning with citizen identification systems.

Main Methods:

• Utilized a multimodal approach combining gait, face, and iris recognition.
• Employed deep convolutional neural networks (DCNNs) pretrained for individual prediction.
• Implemented the system on a cloud server and integrated with Aadhar/SSN systems.

Main Results:

• Achieved a 94% accuracy rate in identifying individuals in public spaces.
• Demonstrated higher accuracy compared to existing public space person identification systems.
• Showcased improved precision and potential for immediate life-saving assistance through system integration.

Conclusions:

• The proposed multimodal secure identification system offers a promising solution for public security threats.
• Deep learning techniques enhance the precision of person identification.
• Potential applications include accident, theft, and intruder identification in public spaces.