Inspiration
The rise of consumer drones has made it easier than ever for unauthorized devices to enter restricted spaces. We noticed a growing problem: most drone-detection systems rely on expensive radar or camera setups, which struggle in low visibility or complex environments.
What inspired us was a simple idea: Drones make a unique sound — what if we teach AI to hear it? This was the question that started our entire journey.
What it does
We were determined to develop a Live Dashboard capable of detecting various drones in noisy outdoor environments to act as support of existing RF and visual detector systems , allowing the user to monitor environments where cameras and RF detection is not practical.
How we built it
Using the CNN model
Challenges we ran into
CNN model failure due to incorrect dataset feed. Dashboard failed to run on the new Model created
Accomplishments that we're proud of
We trained our first machine learning Audio layer 3 CNN model
What we learned
We can accomplish anything with team work
What's next for Ultra Drone Detectors UDD
Improving the AI detection
Log in or sign up for Devpost to join the conversation.