protego

Protego develops audio classification models for surveillance. Audio is position-independent and demands less computational power than video. The model analyzes raw audio data without complex feature extraction, enabling deployment on affordable hardware.

The initial convolutional network attained 98.94% test accuracy in detecting screams from 44.1 kHz raw audio.

Protego focuses on developing audio classification models for edge devices like smartphones, PCs, surveillance cameras, and drones. An edge-first approach enhances privacy, leverages existing computational power, and provides reliable and fast detection.

I'm Raoul Scalise 24, Italian born, now in Lisbon. Led the 11th course at the Italian Military Aeronautical School in 2016. Computer science graduate. 4+ years in mobile development and reverse engineering.

To learn more or see a demo, contact me at scaliseraoul00@gmail.com
I am currently seeking pre-seed funding and excited for possible partnerships.