A Hybrid Deep Learning Model for Robust Biometric Authentication from Low-Frame-Rate PPG Signals

Photoplethysmography (PPG) signals, which measure changes in blood volume in the skin using light, have recently gained attention in biometric authentication because of their non-invasive acquisition, inherent liveness detection, and suitability for low-cost wearable devices. However, PPG signal quality is challenged by motion artifacts, illumination changes, and inter-subject physiological variability, making robust feature extraction and classification crucial. This study proposes a lightweight and cost-effective biometric authentication framework based on PPG signals extracted from low-frame-rate fingertip videos. The CFIHSR dataset, comprising PPG recordings from 46 subjects at a sampling rate of 14 Hz, is employed for evaluation. The raw PPG signals undergo a standard preprocessing pipeline involving baseline drift removal, motion artifact suppression using Principal Component Analysis (PCA), bandpass filtering, Fourier-based resampling, and amplitude normalization. To generate robust representations, each one-dimensional PPG segment is converted into a two-dimensional time-frequency scalogram via the Continuous Wavelet Transform (CWT), effectively capturing transient cardiovascular dynamics. We developed a hybrid deep learning model, termed CVT-ConvMixer-LSTM, by combining spatial features from the Convolutional Vision Transformer (CVT) and ConvMixer branches with temporal features from a Long Short-Term Memory network (LSTM). The experimental results on 46 subjects demonstrate an authentication accuracy of 98%, validating the robustness of the model to noise and variability between subjects. Due to its efficiency, scalability, and inherent liveness detection capability, the proposed system is well-suited for real-world mobile and embedded biometric security applications.

• Publication year

  2025

• Venue

  arXiv.org

• Publication date

  2025-11-06

• Fields of study

  Computer Science, Engineering

• Identifiers
  DOI 10.48550/arXiv.2511.04037arXiv 2511.04037
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Descriptor: Contactless Fingerprint Image Streams and Heart Rate (CFISHR)

  2025cited by this paper
• User Authentication and Vital Signs Extraction from Low-Frame-Rate and Monochrome No-Contact Fingerprint Captures

  2024cited by this paper
• A Secure, Flexible, and PPG-Based Biometric Scheme for Healthy IoT Using Homomorphic Random Forest

  2024influential reference
• Multimodal biometric authentication method by federated learning

  2023cited by this paper
• Ensemble Siamese Network (ESN) Using ECG Signals for Human Authentication in Smart Healthcare System

  2023cited by this paper
• Photoplethysmogram Biometric Authentication Using a 1D Siamese Network

  2023influential reference
• Influencing brain waves by evoked potentials as biometric approach: taking stock of the last six years of research

  2023cited by this paper
• On-Device Deep Learning for Mobile and Wearable Sensing Applications: A Review

  2023cited by this paper
• A PPG-based biometric system using fuzzy min-max model

  2023cited by this paper
• Physical Activity Recognition Based on Deep Learning Using Photoplethysmography and Wearable Inertial Sensors

  2023cited by this paper
• Classification of Post-COVID-19 Emotions with Residual-Based Separable Convolution Networks and EEG Signals

  2023cited by this paper
• Dual-domain and Multiscale Fusion Deep Neural Network for PPG Biometric Recognition

  2023influential reference
• A Novel PPG-Based Biometric Authentication System Using a Hybrid CVT-ConvMixer Architecture with Dense and Self-Attention Layers

  2023cited by this paper
• Challenges and Opportunities of Biometric User Authentication in the Age of IoT: A Survey

  2023cited by this paper
• Reliability of pulse photoplethysmography sensors: Coverage using different setups and body locations

  2022cited by this paper
• Novel Robust Photoplethysmogram-Based Authentication

  2022influential reference
• An Intelligent Multimodal Biometric Authentication Model for Personalised Healthcare Services

  2022influential reference
• Medical image segmentation using deep semantic-based methods: A review of techniques, applications and emerging trends

  2022cited by this paper
• User Authentication Recognition Process Using Long Short-Term Memory Model

  2022influential reference
• Time-Series Forecasting: Extreme Gradient Boosting Implementation in Smartphone Photoplethysmography Signals for Biometric Authentication Processes

  2022influential reference
• Variation-Stable Fusion for PPG-Based Biometric System

  2021cited by this paper
• Self-Attention Bi-LSTM Networks for Radar Signal Modulation Recognition

  2021cited by this paper
• A Review of Wearable Multi-Wavelength Photoplethysmography

  2021cited by this paper
• Evaluation of the Time Stability and Uniqueness in PPG-Based Biometric System

  2021influential reference
• Biosignal classification for human identification based on convolutional neural networks

  2021cited by this paper
• CvT: Introducing Convolutions to Vision Transformers

  2021cited by this paper
• A review of machine learning techniques in photoplethysmography for the non-invasive cuff-less measurement of blood pressure

  2020influential reference
• Deep PPG: Large-Scale Heart Rate Estimation with Convolutional Neural Networks

  2019cited by this paper
• CorNET: Deep Learning Framework for PPG-Based Heart Rate Estimation and Biometric Identification in Ambulant Environment

  2019cited by this paper
• An optimal filter for short photoplethysmogram signals

  2018cited by this paper
• END-to-END Photopleth YsmographY (PPG) Based Biometric Authentication by Using Convolutional Neural Networks

  2018cited by this paper
• Modeling Temporal Structure with LSTM for Online Action Detection

  2018cited by this paper
• TS-LSTM and Temporal-Inception: Exploiting Spatiotemporal Dynamics for Activity Recognition

  2017cited by this paper
• CHAPTER 4 – Time Meets Frequency

  2009cited by this paper
• Photoplethysmography and its application in clinical physiological measurement

  2007cited by this paper
• Interpolation and extrapolation using a high-resolution discrete Fourier transform

  1998cited by this paper

• No citing papers are available for this paper.