AI and Deep Learning in Biometric Security: Trends, Potential, and Challenges - AI and Deep Learning in Biometric Security: Trends, Potential, and Challenges -

Deep Learning-Based Hyperspectral Multimodal Biometric Authentication System Using Palmprint and Dorsal Hand Vein INTRODUCTION DEVICE DESIGN SYSTEM IMPLEMENTATION ROI Extraction Hyperspectral Palmprint ROI Extraction Hyperspectral Dorsal Hand Vein ROI Extraction Feature Extraction Feature Fusion and Matching EXPERIMENTS Multimodal Hyperspectral Palmprint and Dorsal Hand Vein Dataset Optimal Pattern and Band Selection Multimodal Identification Multimodal Verification Computational Complexity Analysis CONCLUSIONS ACKNOWLEDGEMENTS REFERENCES Cancelable Biometrics for Template Protection INTRODUCTION TEMPLATE PROTECTION ROLE OF DEEP LEARNING APPROACHES IN BIOMETRICS RELATED WORK: TEMPLATE PROTECTION Biometric Encryption Biometric Cryptosystems Cancelable Biometrics Deep Learning-Based Cancelable Techniques Deep Learning versus Non-deep Learning Cancelable Techniques PERFORMANCE MEASURES AND DATASETS IN CANCELABLE BIOMETRICS Performance Measures for Non-invertibility Analysis Performance Measures for Unlinkability Analysis Performance Measures for System Usability Analysis Performance Measures for Revocability Analysis Databases Used in Cancelable Biometrics COMPARATIVE PERFORMANCE ANALYSIS: CANCELABLE BIOMETRICS CONCLUSIONS AND FUTURE PROSPECTIVE OF DEEP LEARNING IN BIOMETRICS REFERENCES On Training Generative Adversarial Network for Enhancement of Latent Fingerprints INTRODUCTION RELATED WORK PROPOSED ALGORITHM Problem Formulation and Objective Function Training Data Preparation Network Architecture and Training Details PERFORMANCE EVALUATION Databases and Tools Used Evaluation Criteria RESULTS AND ANALYSIS CHALLENGES OBSERVED CONCLUSIONS ACKNOWLEDGEMENTS REFERENCES DeepFake Face Video Detection Using Hybrid Deep Residual Networks and LSTM Architecture INTRODUCTION RELATED WORK Categories of Face Manipulations DeepFakes Detection PROPOSED DEEPFAKE VIDEOS DETECTION FRAMEWORK Convolutional Neural Networks (CNNs)Long Short-Term Memory (LSTM)Residual Neural Network (ResNet)EXPERIMENTS Datasets Figures of Merit Experimental Protocol Experimental Results CHALLENGES AND FUTURE RESEARCH DIRECTIONS CONCLUSIONS NOTES REFERENCES Multi-spectral Short- Wave Infrared Sensors and Convolutional Neural Networks for Biometric Presentation Attack Detection INTRODUCTION DEFINITIONS RELATED WORKS PROPOSED PAD METHOD Hardware: Multi-Spectral SWIR Sensor Software: Multi-Spectral Convolutional Neural Networks Multi-Spectral Samples Pre-Processing CNN Models Score Level Fusion EXPERIMENTAL SETUP Database Evaluation Metrics Experimental Protocol EXPERIMENTAL EVALUATION Baseline: Handcrafted RGB Conversion Input Pre-Processing Optimisation Final Fused System CONCLUSIONS AND FUTURE RESEARCH ACKNOWLEDGEMENTS REFERENCES Al-Based Approach for Person Identification Using ECG Biometric INTRODUCTION ECG AND RELATED WORK METHODOLOGY ADOPTED Feature Extraction CLASSIFIER Artificial Neural Network (ANN)Support Vector Machine (SVM)EXPERIMENTS AND RESULTS CONCLUSIONS REFERENCES Cancelable Biometric Systems from Research to Reality INTRODUCTION CANCELABLE BIOMETRIC SYSTEMS: INTRODUCTION AND REVIEW Conventional Template Transformation Techniques Role of Deep Learning in Biometrics and Need for Privacy Neutral Network-Based Template Transformation Techniques EXPERIMENTAL REPORTING REAL-LIFE CHALLENGES FOR APPLICATIONS OF CANCELABLE BIOMETRIC SYSTEMS CONCLUSIONS AND FORESIGHTS REFERENCES Gender Classification under Eyeglass Occluded Ocular Region INTRODUCTION Our Contributions RELATED WORKS Visible Spectrum Near-Infra-Red Spectrum Visible and Near-Infra-Red Spectrum Multi-Spectral Imaging DATABASE PROPOSED METHOD Spectral Bands Selection Feature Extraction Classification EXPERIMENTS AND RESULTS Experimental Evaluation Protocol Evaluation 1: Without-Glass v/s Without-Glass Individual Band Comparison Fused Band Comparison Evaluation 2: Without-Glass v/s With-Glass Individual Band Comparison Fused Band Comparison CONCLUSIONS ACKNOWLEDGEMENT REFERENCES Investigation of the Fingernail Plate for Biometric Authentication using Deep Neural Networks INTRODUCTION Motivation and Scope of Present Work RELATED WORK SAMPLE ACQUISITION AND ROI EXTRACTION Sample Acquisition ROI Extraction FEATURE EXTRACTION Transfer Learning using AlexNet Transfer Learning using ResNet-18 Transfer Learning using DenseNet-201 MULTIMODAL SYSTEM DESIGN Score-Level Fusion Rank-Level Fusion Logistic Regression Method Mixed Group Rank Inverse Rank Position Nonlinear Weighted Methods EXPERIMENTS, RESULTS, AND ANALYSES Performance of Fingernail Plates in Verification Systems Performance of Fingernail Plates in Unimodal Verification Systems Performance of Fingernail Plates in Multimodal Verification Systems Performance of Fingernail Plates in Identification Systems Performance of Fingernail Plates in Unimodal Identification Systems Э.6.2.2 Performance of Fingernail Plates in Multimodal Identification Systems CHALLENGES AND SCOPE OF FINGERNAIL PLATES IN BIOMETRICS CONCLUSIONS AND FUTURE SCOPE REFERENCES Fraud Attack Detection in Remote Verification Systems for Non-enrolled Users INTRODUCTION RELATED WORK Remote Authentication Framework Using Biometrics Image Manipulation and Deep Learning Techniques FAKE ID CARD DETECTION FOR NON-ENROLLED USERS Databases Hand-Crafted Feature Extraction (BSIF, uLBP, and HED)Automatic Feature Extraction (CNN)EXPERIMENTS AND RESULTS Feature Extraction Classification Classification Using CNN Algorithms Small-VGG Trained from Scratch Pre-trained VGG16 Model and Bottleneck Pre-trained VGG16 Model and Fine-Tuning CONCLUSIONS ACKNOWLEDGEMENT REFERENCES Indexing on Biometric Databases INTRODUCTION INDEXING FACIAL IMAGES Predictive Hash Code INDEXING FINGERPRINT IMAGES Coaxial Gaussian Track Code INDEXING FINGER-KNUCKLE PRINT DATABASE Boosted Geometric Hashing INDEXING IRIS IMAGES Indexing of Iris Database Based on Local Features INDEXING SIGNATURE IMAGES KD-Tree-Based Signature Database Indexing CONCLUSION REFERENCES Iris Segmentation in the Wild Using Encoder-Decoder-Based Deep Learning Techniques INTRODUCTION DEEP LEARNING FOR SEGMENTATION RELATED WORK Non-Deep Learning-Based Methodologies Deep Learning-Based Methodologies DATA SETS AND EVALUATION METRICS Data sets. CASIA Performance Metrics EXPERIMENTATION CHALLENGES IDENTIFIED AND FURTHER DIRECTION CONCLUSION ACKNOWLEDGEMENTS REFERENCES PPG-Based Biometric Recognition INTRODUCTION PHOTOPLETHYSMOGRAM (PPG)LITERATURE REVIEW MULTI-FEATURE APPROACH FOR PPG BIOMETRIC CLASSIFICATION EXPERIMENTS AND RESULTS CONCLUSIONS REFERENCES Current Trends of Machine Learning Techniques in Biometrics and its Applications INTRODUCTION Biometric Systems Brain Stroke Face Recognition Motivation to Machine Learning Techniques RELATED WORK Review on Brain Stroke Review on Face Recognition Brain Stroke Prediction System Image Acquisition Pre-processing Feature Extraction Classification Using Machine Leaning Algorithms Decision Tree Artificial Neural Network Support Vector Machine Deep Learning with CNN Construction of Convolutional Neural Network Face-Recognition System DISCUSSION AND RESULTS Performance of Brain Stroke Performance of Face Recognition FUTURE SCOPE CONCLUSION REFERENCES