Pupil Tracking GPU _v1.0.0

Harnessing GPU acceleration for efficient and accurate pupil tracking in real-time applications.

---

📢 Disclaimer

This software is developed by the University of Murcia and is intended solely for academic and research purposes. It must not be used for commercial or business applications.

🌟 Introduction

Welcome to the Pupil Tracking GPU repository! This project is designed to leverage the power of modern GPUs for high-performance, real-time pupil tracking. Whether you're involved in eye-tracking research, developing advanced human-computer interaction (HCI) systems, or exploring applications in medical diagnostics, this repository offers a robust and adaptable foundation. Our goal is to provide a tool that is both powerful and accessible to the research community.

✨ Features

• 🚀 GPU Acceleration: Utilizes CUDA for significant performance gains, enabling complex computations at high frame rates.
• ⏱️ Real-Time Processing: Achieves low-latency tracking suitable for interactive and time-sensitive applications.
• 🔧 Customizable Algorithms: Implements multiple pupil detection algorithms (ELSE, EXCUSE, and their greedy variants) allowing users to choose based on their specific needs for accuracy and speed.
• 🛠️ Flexible Platform: Supports execution on both GPU and CPU for broader compatibility and testing.
• 📊 Performance Measurement: Integrated option to measure execution time for benchmarking and optimization.
• 📖 Open Source: Developed for the academic and research community, encouraging contributions and collaborative improvement.

⚙️ Installation

Follow these steps to get the Pupil Tracking GPU software up and running:

1. Clone the Repository:

   git clone https://github.com/artuppp/PupilTrackingGPU.git
   cd PupilTrackingGPU

2. Install Dependencies: This project relies on OpenCV.

   sudo apt-get update
   sudo apt-get install libopencv-dev

3. GPU Setup:

   • Ensure you have a CUDA-compatible NVIDIA GPU.
   • Install the latest NVIDIA drivers for your GPU.
   • Install the NVIDIA CUDA Toolkit compatible with your drivers and project requirements.

4. Build the Project: A Makefile is provided for easy compilation.

   make

   This will create the executable in the build/ directory.

🚀 Usage

The main executable pupil_tracking can be run from the command line with several arguments to control its behavior.

Synopsis:

./build/pupil_tracking <image_path> <measure_time> <platform> <num_repetitions> <algorithm>