Multi-Modal Face Anti-Spoofing System Using Deep Learning: Integrating Facial Liveness, Eye  Blink Detection, and Voice Authentication for Robust Biometric Security

Gayathri Jayamohan; Kiran Kumar B

Multi-Modal Face Anti-Spoofing System Using Deep Learning: Integrating Facial Liveness, Eye Blink Detection, and Voice Authentication for Robust Biometric Security

Author(s)	Ms. Gayathri Jayamohan, Dr. Kiran Kumar B
Country	India
Abstract	Biometric authentication systems face an unprecedented surge in sophisticated spoofing attacks that exploit single-modality verification pipelines. This paper proposes MFASNet — a Multi-Modal Face Anti-Spoofing Network — that fuses three independent verification channels: (1) facial liveness detection classifying inputs as live, photograph, video-replay, or 3-D mask attacks; (2) eye blink detection as a physiological liveness cue using a Convolutional Long Short-Term Memory (ConvLSTM) temporal model; and (3) voice anti-spoofing distinguishing genuine human speech from recorded, replayed, and AI-generated audio using a Res2Net acoustic model. A late-fusion decision module aggregates the three scores through a learned meta-classifier, granting access only when all three streams concurrently classify the input as genuine. Experimental evaluation on FaceForensics++, CelebA-Spoof, ASVspoof 2021 LA, and NUAA datasets demonstrates that MFASNet achieves a cross-dataset Equal Error Rate (EER) of 2.14%, Area Under the Curve (AUC) of 99.31%, and Half Total Error Rate (HTER) of 1.97%, outperforming state-of-the-art unimodal and bimodal baselines by a significant margin. The novelty of incorporating temporally-aware eye blink cadence as an auxiliary liveness stream yields a 3.8% absolute reduction in False Acceptance Rate compared to face-only systems. The proposed architecture is lightweight (12.4 M parameters) and inference-efficient (34 ms on an NVIDIA RTX 3060), making it suitable for real-time deployment in access-control environments.
Keywords	Face anti-spoofing; voice anti-spoofing; eye blink detection; multimodal fusion; deep learning; biometric security; presentation attack detection.
Field	Computer > Artificial Intelligence / Simulation / Virtual Reality
Published In	Volume 8, Issue 2, March-April 2026
Published On	2026-03-13

View / Download PDF File

E-ISSN 2582-2160

doi

CrossRef DOI is assigned to each research paper published in our journal.

IJFMR DOI prefix is
10.36948/ijfmr

Downloads

Research Paper Format Copyright Permission Form and Undertaking Form Cover Page Vol 8 Isu 2 Cover Page Vol 8 Isu 1 Cover Page Vol 7 Isu 6

All research papers published on this website are licensed under Creative Commons Attribution-ShareAlike 4.0 International License, and all rights belong to their respective authors/researchers.

CC-BY-SA

About IJFMR Fees & Payment Current Issue Publication Archive	Submit Research Paper Track Submission Status Publication Guidelines Publication Ethics Peer Review & Plagiarism	Join as a Reviewer Editors & Reviewers Reviewer Referral Program Get Reviewer Membership Certi.	Website/Journal Policies Usage Policy Content Policies Privacy Policy

Contact Us		+91-9687-828-838	editor@ijfmr.com

International Journal For Multidisciplinary Research

A Widely Indexed Open Access Peer Reviewed Multidisciplinary Bi-monthly Scholarly International Journal

Multi-Modal Face Anti-Spoofing System Using Deep Learning: Integrating Facial Liveness, Eye Blink Detection, and Voice Authentication for Robust Biometric Security

Share this