A Review of YOLOv7-based Multimodal Biometric Frameworks: Scalability, Robustness, and Identification Performance

Description

Page: 36-40

Sheetal Shevkari¹, Kelapati², and Sharmila More³ (MIT Arts, Commerce & Science College, Alandi, Pune, Maharashtra^1,3, Shri Jagdishprasad Jhabarmal Tibrewala University, Rajasthan²)

This review explores the integration of YOLOv7, a high-performance object detection model, into multimodal biometric systems to enhance scalability and robustness in real-time applications, particularly in high-security sectors like airports and banking. YOLOv7’s speed and accuracy make it a promising candidate for feature extraction and dataset management, offering advantages in biometric identification systems. The review discusses the benefits of fusion techniques that combine data at various levels to improve accuracy, as well as the challenges associated with YOLOv7’s use in biometric tasks. Future research directions include the development of hybrid fusion methods, privacy-preserving techniques, and the application of edge computing for large-scale deployment. This paper aims to provide a comprehensive understanding of how YOLOv7 can advance multimodal biometric systems, addressing emerging challenges in identification and security.