Researchers at the University of Tokyo have developed LatentCSI, a groundbreaking technology that employs Stable Diffusion 3 and Wi-Fi Channel State Information (CSI) to create real-time images of rooms and their occupants. This innovative method utilizes invisible Wi-Fi signal data to capture strikingly accurate visuals, demonstrating how much information can be extracted from Wi-Fi signals. Although highly precise imagery requires prior photographs of the target environments, the advancement highlights a significant leap toward next-gen surveillance without traditional cameras. Previously, Wi-Fi CSI enabled motion tracking, but LatentCSI enhances this by converting signal data into visually coherent images. While concerns about privacy arise, the technology mirrors existing Wi-Fi motion sensing capabilities. LatentCSI may be particularly useful in settings devoid of cameras, serving private businesses and government agencies. This reflects the evolving landscape of AI-driven surveillance, underscoring both its potential benefits and ethical considerations.
Source link
