Hatem Zehir, PhD
Researcher in Biometrics · Deep Learning · Signal Processing · Embedded AI
Laboratory of Study and Research in Instrumentation and Communication of Annaba (LERICA), Badji Mokhtar-Annaba University, P.O. BOX 12, 23000, Annaba, Algeria
Download Full CV (PDF)ECG Biometrics
Facial Recognition
Multimodal Biometrics
TinyML & Edge AI
Deep Learning
Signal Processing
Computer Vision
PhD in Security & Biometrics
Badji Mokhtar — Annaba University, Algeria
2022 – 2025
Thesis: Development of a Hybrid Multimodal Biometric System
MSc in Instrumentation
Badji Mokhtar — Annaba University, Algeria
2019 – 2021
Dissertation: Design and Implementation of a Smart Contactless Thermometer with Facial Recognition
BSc in Electronics
Badji Mokhtar — Annaba University, Algeria
2016 – 2019
Honorary Research Fellow
Beijing Institute of Technology
Developing Python and MATLAB algorithms to filter and structure fluid dynamics data (noise removal, physical reflection correction) for integration with deep learning models. Experimental data sourced from BIT laboratories, methodology grounded in NASA-developed Particle Image Velocimetry (PIV) processing frameworks.
Doctoral Researcher
LERICA Laboratory, Badji Mokhtar University
Conducted PhD research on deep learning-based multimodal biometric systems. Developed and evaluated ECG-based identification pipelines using CNN, LSTM, GRU, and BiLSTM architectures. Designed score-level fusion frameworks combining ECG and voice modalities achieving 100% identification accuracy. Published 10+ journal articles and 7 conference papers in international venues.
PRFU Project Research Member
LERICA Laboratory, Badji Mokhtar University
Contributing to the national research project “Multidimensional Signal Processing: Applications in Biometrics.” Responsible for developing feature extraction algorithms (EMD, HHT, spectrograms), designing deep learning evaluation frameworks, and implementing TinyML solutions for embedded biometric devices on ESP32 platform.
-
Unifying Heartbeats and Vocal Waves: An Approach to Multimodal Biometric Identification At the Score Level
-
Empirical mode decomposition-based biometric identification using GRU and LSTM deep neural networks on ECG signals
-
An ECG Biometric System Based on Empirical Mode Decomposition and Hilbert-Huang Transform for Improved Feature Extraction
Programming
Python, MATLAB, C, C++, Basic PHP
ML/DL Frameworks
TensorFlow, Keras, PyTorch, scikit-learn, Edge Impulse
Embedded & IoT
ESP32, Arduino, Raspberry Pi, ARM Cortex-M, TinyML, Thinger.io, MQTT
Signal & Image Processing
OpenCV, FFT, EMD, HHT, MFCC, Spectrogram analysis, Digital filtering
Tools & Platforms
Git, Linux, LaTeX, Jupyter Notebook, MATLAB Simulink, PCB Design (KiCad/Eagle)
Research & Writing
Academic writing, Experimental design, Dataset collection, Peer review
🇩🇿
Arabic
🇬🇧
English
🇫🇷
French
Visit dedicated sections of my website for comprehensive information: