Recent Papers

The primary purpose of this research is to develop a secure and intelligent banking authentication system that integrates face recognition with liveness detection to prevent identity fraud and unauthorized access in digital financial services. Traditional authentication methods such as passwords, PINs, and tokens are prone to phishing and cyberattacks, while conventional facial recognition systems can be easily deceived by printed photos, replayed videos, or 3D mask attacks. To address these limitations, this study proposes a Machine Learningbased approach utilizing Image Processing and Deep Learning techniques for robust identity verification. The system is designed using a Convolutional Neural Network (CNN) for accurate facial feature extraction and recognition, combined with a liveness detection module that analyzes natural human behaviors such as eye blinking, lip movement, and subtle facial micro-expressions to differentiate between live users and spoofing attempts. The model is trained and validated using publicly available datasets such as ORL, OULU, and CASIA, ensuring adaptability to diverse illumination, pose, and texture conditions. The implementation, developed in Python using TensorFlow, OpenCV, and Keras with a MySQL database, was tested for accuracy, latency, and resilience to spoofing attacks. Experimental analysis demonstrated that the proposed system achieved superior performance with over 98% recognition accuracy, a low false acceptance rate (FAR), and real-time response suitable for online and mobile banking environments. The results indicate that combining CNN-based recognition with motion-aware liveness verification significantly enhances authentication reliability compared to traditional methods. The study concludes that the proposed system provides a cost-effective, efficient, and user-friendly solution that strengthens banking security infrastructure. Moreover, its modular design allows seamless integration with existing banking platforms while ensuring data privacy and regulatory compliance. Overall, the research establishes that fusing machine learning and image processing techniques can revolutionize biometric security, setting a foundation for next-generation banking systems capable of resisting evolving digital spoofing and fraud attempts.

Copyright © 2025 Jog Sakshi Vinayak. This is an open access article distributed under the Creative Commons Attribution License.