Oosto branching out with cloud-to-edge pivot for facial recognition tech
Facial recognition company Oosto has announced plans to move its flagship Vision AI software from embedded apps in edge devices to edge-to-cloud algorithms.
Distributed security, surveillance and safety uses made possible by this development include skeletal modeling, anomaly detection, time series tracking, behavior analysis and pattern tracking, according to the company.
Oosto says its neural networks will run on onboard cameras while leveraging GPUs in near-edge devices (also called on-premise edge) and delivering metadata to the cloud for analytics processing. The company claims its algorithms are seven times more efficient in terms of watts-per-video stream than competitors, noting that one of the world’s largest commercial security system integrators has embedded Oosto inside the company’s AI cameras and video monitoring system.
Oosto says its software now can create a composite of a person from their face and body attributes and track that person live from camera to camera. For example, it can see if employees are consistently wearing required protective gear as they move throughout the premises.
Addressing large-scale, distributed cloud or mixed environments could move the company beyond security and into manufacturing and logistics, schools, health care and transportation, the company says.
Complicating the strategic shift is Oosto’s decision to shake up the organization chart. The company has axed 10 of its 95 employees.
Amdocs, Microsoft collaborate on developing private networks and edge computing solutions
device edge | edge AI | facial recognition | neural network | Oosto | video surveillance