Engineering Animation
Vol. 40 No. 05 (2025): Proceedings of the Faculty of Technical Sciences
USE OF COMPUTER VISION METHODS FOR USER INTERFACE CONTROL OF INTERACTIVE SOFTWARE
Abstract
This paper investigates the application of computer vision to control the user interface, specifically the input devices of a computer, and management of a game created in Unreal Engine. Python, MediaPipe, and PyAutoGUI libraries were used. The controls are implemented in real-time through communication between a Python script and the game engine.
References
- [1] Dragan Micić, Tončić Danilo – "Lokalizacija na osnovu markera", https://esveske.github.io/pdf/2017/PFE1701.pdf
- [2] Python Documentation – https://docs.python.org/3/tutorial/
- [3] Pajankar, Ashwin – "Naučite Python 3", Agencija Eho, 2022
- [4] PyAutoGUI Documentation – https://pyautogui.readthedocs.io/en/latest/
- [5] Dattanikaran – "Get Started with PyAutoGUI", Medium
- [6] Google – MediaPipe Solutions Guide, https://ai.google.dev/edge/mediapipe/solutions/guide
- [7] MediaPipe Hands Documentation – https://mediapipe.readthedocs.io/en/latest/solutions/hands.html
- [8] MediaPipe Face Mesh Documentation – https://mediapipe.readthedocs.io/en/latest/solutions/face_mesh.html
- [9] Ushaw G., Alanazi F. – "Improving Detection of DeepFakes through Facial Region Analysis", ResearchGate
- [10] RisingStack – "Node Hero – Getting Started With Node.js Tutorial", https://blog.risingstack.com/node-hero-tutorial-getting-started-with-node-js/
- [11] Bogomyakov E. – "How to connect Unreal Engine 4 to local server via WebSocket using blueprints?", Medium