DiveGuard - smart system for underwater navigation and communication

DiveGuard is an innovative system developed by me with elements of artificial intelligence that ensures safe and efficient diver immersion. It combines computer vision technologies, gesture recognition, voice control, speech synthesis, and natural language processing.

The system can:
- Detect dangerous objects (e.g., sharks);
- Determine the diver’s health status and environmental parameters;
- Respond to user gestures (e.g., "SOS", "All clear", "I am ascending", etc.);
- Display information on a Head-Up Display interface;
- Communicate with the user via voice and text without manual control.

Technologies used:
Python, YOLO, PyQt5, MediaPipe, Roboflow, SpeechRecognition, pyttsx3, FLAN-T5 XL.

Result:
DiveGuard significantly improves diving safety, suitable for ecological, scientific, or rescue underwater operations. The program has been tested in simulations and is ready for integration into hardware solutions.

Achievements:
This project has been presented at several international competitions, where it received recognition:
- GENIUS Olympiad (USA) — Silver Medal
- Mostratec (Brazil) — Official Participation
- KIDE Expo (Taiwan) — Bronze Medal + Special Prize
- INTARG (Poland) — Silver Medal
- Sakura Science Exchange Program (Japan) — Participation in exchange program