Logo

Projects

The deployment of computer vision solutions across various hardware platforms.

Hardware deployment evolution

Microcontroller

STM32

Embedded fingerprint recognition

Memory-constrained image processing and matching in C.

Single-board computer

Raspberry Pi

Object detection and sorting

Lightweight real-time vision with OpenCV and compact models.

Accelerated edge AI

NVIDIA Jetson

Industrial edge monitoring

GPU-accelerated deployment for Edge AI and IoT systems.

Server infrastructure

GPU Server

Multi-camera vision platform

Concurrent real-time detection, tracking, and other vision tasks.

Computer Vision

Image processing, computer vision deep learning, real-time applications, edge deployment.

Multi-camera HydraNet

Smart City Vision System

Real-time multi-camera detection, tracking, segmentation, classification, and face recognition with a shared-backbone architecture.

  • PyTorch
  • OpenCV
  • ONNX
  • TensorRT
View case study →
Coffee sorting system

Coffee Bean Sorting Machine

End-to-end image acquisition, enhancement, segmentation, classification, and embedded optimization for non-uniform lighting conditions.

  • Jetson Nano
  • Raspberry Pi
  • OpenCV
  • CIELAB
Case study coming soon
STM32 fingerprint pipeline

Embedded Fingerprint Recognition

Raw fingerprint acquisition through SPI, enhancement using STFT and Gabor filtering, ridge analysis, feature extraction, and efficient matching.

  • C
  • STM32
  • SPI
  • Gabor
Case study coming soon
Raspberry Pi object detection

Real-Time Object Detection on Raspberry Pi

A lightweight OpenCV and MobileNet application for real-time object detection on Raspberry Pi 3.

  • Raspberry Pi 3
  • OpenCV
  • MobileNet
Case study coming soon
OCR workflow

Name Card Recognition

An image-processing and OCR workflow that extracts structured information such as names, phone numbers, and addresses.

  • Python
  • OpenCV
  • PyTesseract
  • Regex
Case study coming soon

Explainable AI

Pre-model, during-model, and post-model explainability for industrial applications.

Jetson edge monitoring

Edge IoT Monitoring

Industrial edge monitoring on Jetson Orin Nano with federated learning integration.

  • Jetson Orin Nano
  • Edge AI
  • IoT
  • Federated Learning
Case study coming soon
Multimodal explanation framework

Post-Hoc Multimodal Explainability

A framework for interpreting deep learning models that combine image, text, and numerical sensor data.

  • XAI
  • Multimodal AI
  • Industrial AI
Research case study coming soon
Interpretable prognostics

Intrinsically Explainable Health Monitoring

An intrinsically explainable framework for system-level health monitoring and prognostics.

  • Interpretable AI
  • Prognostics
  • Health Monitoring
Research case study coming soon
Data-quality explainability

Pre-Model Explainability

Methods for analyzing data quality before model training and supporting AI model development for industrial prognostics.

  • Data Quality
  • Pre-Model XAI
  • Industrial AI
Research case study coming soon

Other Projects

Signal processing and related projects.

Neural signal analysis

Neural Recording & EEG Event Detection

A MATLAB interface for neural signal analysis and visualization, with FIR filtering and EEG event-detection algorithms.

  • MATLAB
  • FIR Filters
  • EEG
  • Signal Processing
Project summary coming soon

Interested in collaborating?

I am open to applied research and engineering collaborations in edge computer vision, explainable AI, and industrial AI.

Contact me