Tomaz Zlindra
chip_circuit_processor_140251_1920x1080
Computer Vision Project

Electrical Engineering Student;

Built a YOLO-based system that detects GUI buttons from screenshots, trained on a personal NVIDIA GPU, and designed for future integration with OCR and voice commands to support accessibility.

Screenshot 2025-09-03 205659

Personal, Vancouver, BC, Canada

April 2025

Skills: python, YOLO, Machine Learning, coco

  • Collected screenshots and real-time GUI imagery to build a dataset of button targets, annotated and preprocessed for training.
  • Trained a custom YOLO model using a personal NVIDIA GPU, enabling efficient local experimentation, faster iterations, and fine-tuning for button detection tasks.
  • Applied YOLO machine learning for real-time button recognition, logging, and automated mouse control, demonstrating practical ML integration with computer vision and system-level interaction.
  • Designed with future extensions in mind, where YOLO detection could work alongside OCR and voice commands to aid visually impaired users in navigating computer interfaces.