AI & Computer Vision R&D Engineer : turning industrial and biomedical signals into explainable visual intelligence
| 🎓 ENSTA Institut Polytechnique de Paris |
🔬 R&D Intern IFP Énergies Nouvelles Lyon |
📄 2 IEEE Papers Peer-reviewed ICABME · ETECOM |
🚀 Available October 2026 R&D positions |
I work at the intersection of signal processing, deep learning, and computer vision, with a focus on making models both accurate and explainable. My current research at IFPEN encodes multivariate industrial time series as images so that convolutional architectures, and their explainability tools, can detect and localize anomalies in complex physical systems. Before that, I applied the same signal-to-model philosophy to biomedical data: ECG-based atrial fibrillation detection (two IEEE publications), depth-camera pose estimation for medical robotics, and EEG-based brain–computer interfaces. What drives me is research that survives contact with real sensors, real noise, and real industrial constraints.
Pipeline
Multivariate signals → Image encoding (GADF · GASF · MTF) → Convolutional autoencoder
→ Anomaly score → Explainability (Grad-CAM · LRP) → Sensor-level localization
- Signal-to-image encoding of multivariate sensor data using Gramian Angular Fields (GADF, GASF) and Markov Transition Fields (MTF)
- Convolutional autoencoders for fully unsupervised fault detection
- Sensor-level fault localization through Grad-CAM and Layer-wise Relevance Propagation (LRP)
- Validated on the Tennessee Eastman Process benchmark
- Complementary validation on real industrial pilot data
| Period | Role | Organization |
|---|---|---|
| 🟠 Mar. 2026 – Present | R&D Intern — Computer Vision & Anomaly Detection | IFP Énergies Nouvelles |
| May 2025 – Sep. 2025 | R&D Intern — Depth Cameras & Pose Estimation | Ivanae Medical / LaTIM |
| Dec. 2023 – Jun. 2024 | AI Engineer — License Plate Detection | RODOK SARL |
| May 2023 – Aug. 2023 | AI Engineer — Cardiac Signal Classification | Together for Chehim |
|
ENSTA · 2025–2026 Explainability of chest X-ray models (CheXNet, PYLON) using Grad-CAM, Grad-CAM++, LRP, and Integrated Gradients for trustworthy medical imaging. |
Personal project · 2024–2025 Transformer U-Net on multimodal MRI, targeting precise segmentation of the WT, TC, and ET tumor subregions. |
|
ENSTA · 2024–2025 Real-time multi-object detection and tracking with a YOLOv10 + DeepSORT pipeline on multi-stream video. |
Université Libanaise · 2024 EEG signal processing, noise reduction, feature extraction, and machine-learning classification for BCI applications. |
-
Evaluating the Impact of EMD-Derived Intrinsic Mode Functions on Atrial Fibrillation Detection ICABME 2025 — DOI: 10.1109/ICABME66883.2025.11211811
-
A Data-Augmented Deep Hybrid Model for Atrial Fibrillation Detection ETECOM 2025 — DOI: 10.1109/ETECOM66111.2025.11319010
Full technical toolbox
| Domain | Technologies |
|---|---|
| AI & Computer Vision | PyTorch · TensorFlow · OpenCV · MediaPipe · YOLO · scikit-learn · autoencoders · Transformers · explainability methods (Grad-CAM, LRP, Integrated Gradients) |
| Data & Scientific Computing | NumPy · Pandas · Matplotlib · Seaborn · signal processing · time-series analysis |
| Programming Languages | Python · C · C++ · Java · JavaScript · LaTeX |
| Engineering Tools | Git · Linux · Docker · VS Code · Google Colab |
| Period | Degree | Institution |
|---|---|---|
| 2024 – 2026 | Engineering Diploma — Digital Systems Design | ENSTA – Institut Polytechnique de Paris |
| 2021 – 2024 | Bachelor's Degree — Networks & Telecommunications | Université Libanaise |
- Arabic — Native
- English — TOEIC 940
- French — B2
I am open to R&D engineering positions in AI and Computer Vision (from October 2026), as well as collaborations and technical discussions around:
computer vision · signal processing · applied deep learning · industrial AI · explainable AI