Alessandro Capotondi's team focuses on optimizing AI models for edge devices using GPUs and FPGAs. He explains that their work primarily revolves around fine-tuning artificial intelligence models to ensure they operate efficiently on edge devices, such as smart cameras.
The goal is to deploy these technologies within the context of smart cities, leveraging the capabilities of smart cameras to autonomously detect vehicles, pedestrians, and other road users. This distributed system not only improves road safety but also provides real-time information to connected vehicles, optimizing traffic management and safety.
The impact of the dAIEDGE project extends beyond technology. Alessandro notes that the work package is led by Hypertension, a startup that emerged from his university, demonstrating how research is translating into real-world products. This collaborative model has a direct economic impact, as startups bring innovations to market.
Furthermore, as part of his role at the university, Alessandro emphasizes the educational impact of the project, as they are training new generations of students to specialize in these emerging technologies, thereby enhancing their future job prospects.
You can watch the full interview with Alessandro Capotondi of Unimore here: https://www.youtube.com/watch?v=MEFpqMy15WM&ab_channel=dAIEDGE