Dr.Anusha M S
Assistant Professor
About
Dr. Anusha M.S. is an Assistant Professor in the School of Computer Science and Engineering at RV University. She is passionate about fostering an engaging learning environment that encourages curiosity, critical thinking, innovation, and lifelong learning. With a strong commitment to academic excellence, she believes in bridging theoretical concepts with practical applications to prepare students for the evolving technological landscape.
Her research interests include Artificial Intelligence, Machine Learning, Deep Learning, Computer Vision, Biometrics, Pattern Recognition, and Image Processing, with a particular focus on fingerprint recognition and fingerprint liveness detection. She has published research in reputed national and international journals and conferences and continues to contribute to advancing knowledge in intelligent computing and biometric security.
As an educator, Dr. Anusha strives to inspire students to think beyond conventional solutions by integrating research into classroom learning. She actively mentors students in projects, research initiatives, and technical activities, encouraging them to develop innovative solutions to real-world challenges. She also promotes professional ethics, collaborative learning, and a research-driven mindset among aspiring engineers.
Her academic philosophy is centered on the belief that education and research together have the power to create meaningful societal impact. At RV University, she looks forward to contributing to academic excellence through quality teaching, impactful research, and interdisciplinary collaboration while nurturing future engineers and researchers capable of driving technological innovation.
This Scopus Q2 journal publication proposes a lightweight Convolutional Neural Network (CNN) architecture for fingerprint spoof detection by integrating gradient-based ridge features with sweat pore characteristics. The proposed framework enhances fingerprint liveness detection by effectively capturing both macro-level ridge patterns and micro-level pore information while maintaining computational efficiency. The study further incorporates visual explainability techniques to improve the transparency and interpretability of the deep learning model, enabling better understanding of the decision-making process. Experimental evaluation demonstrates high detection accuracy and robustness against presentation attacks, making the proposed model suitable for secure biometric authentication systems. The research contributes toward the development of efficient, explainable, and reliable AI-based biometric security solutions.
DOI: 10.53894/ijirss.v8i3.7231
This research presents a feature fusion framework for fingerprint spoof detection by combining gradient-based ridge information with sweat pore features to enhance biometric authentication accuracy. The proposed approach leverages complementary fingerprint characteristics to improve the discrimination between genuine and spoof fingerprints while increasing system robustness against presentation attacks. The study investigates the effectiveness of multimodal feature fusion in improving classification performance and reducing misclassification rates. Experimental results demonstrate that integrating multiple biometric features significantly enhances fingerprint liveness detection compared to conventional single-feature approaches. The work contributes to the advancement of secure biometric authentication systems by providing an effective and reliable solution for fingerprint presentation attack detection using intelligent feature fusion techniques.
This research investigates fingerprint liveness detection techniques aimed at preventing spoof attacks on biometric authentication systems. The study evaluates both hardware-based and software-based approaches for distinguishing genuine fingerprints from artificial or fake fingerprints. Various fingerprint characteristics and liveness indicators are analyzed to improve the robustness of biometric security systems while minimizing false acceptance of spoof fingerprints. The work highlights the strengths and limitations of different liveness detection methods and emphasizes the importance of combining hardware and software solutions to achieve reliable authentication. The research contributes to enhancing the security of fingerprint-based access control systems and was presented through an IEEE publication, demonstrating the practical significance of biometric anti-spoofing technologies.
Research interests: Artificial Intelligence, Machine Learning, Deep Learning, Computer Vision, Biometrics, Pattern Recognition, and Image Processing.
Specialization: Fingerprint recognition, fingerprint liveness detection, presentation attack detection, and AI-driven biometric security.
Research contributions: Developed intelligent biometric authentication frameworks using CNNs, SVMs, and advanced fingerprint feature analysis for enhanced security.
Publications: Published research in Scopus-indexed (Q2) journals, IEEE conferences, and other peer-reviewed international journals.
Research vision: Committed to advancing trustworthy and explainable AI while mentoring students in research, innovation, and scientific publishing.
Research Interests
Artificial Intelligence (AI)
Machine Learning & Deep Learning
Computer Vision
Biometrics & Biometric Security
Fingerprint Recognition & Liveness Detection
Pattern Recognition
Image Processing
Reviewer – Third IEEE International Conference on Emerging Computation and Information Technologies (ICECIT-2025)
Served as a reviewer for the Third International Conference on Emerging Computation and Information Technologies (ICECIT-2025) at Siddaganga Institute of Technology, evaluating technical papers and supporting the peer-review process.
Dr.Anusha M S
Assistant Professor
Ph.DSchool of Computer Science and Engineering
-
anushams@rvu.edu.in -
Bengaluru
Academics
Admissions
Faculty
Enquire