Real-time object detection on microprocessor (Open Source Project)

Docente

Fabio Salice (mail)

Referente del progetto

Fabio Salice ( mail)

Area di ricerca

Architetture dei sistemi di elaborazione, Intelligenza artificiale e robotica

Keyword (max 3 separate da virgola)

AI, Microntroller

Termine per accettazione progetto

Descrizione (max 500 caratteri)

This project is proposed by Zant. Zant is creating an open-source SDK to simplify and cost-effectively deploy machine learning (ML) models on embedded and edge devices.

The project focuses on optimizing YOLO (You Only Look Once), a state-of-the-art real-time object detection algorithm, for microprocessors. By enhancing YOLO for resource-constrained environments, the initiative enables edge-based AI applications like robotics, surveillance, and IoT, demonstrating how advanced computer vision can run on minimal hardware for low-cost, power-efficient solutions.

Participants will be mentored by project founders, learning microprocessor basics, ML libraries, and open-source collaboration. They will work in global teams, including Carnegie Mellon students, to improve YOLO’s performance on microprocessors using state-of-the-art techniques.

Scroll to Top