Speaker: Dr. Alberto Stochino

Institution: Perceptive, Inc.

Autonomous perception for driverless vehicles has set extremely high requirements on imaging sensors such as lidars, radars, and cameras. Perception processes like clustering, classification, and tracking for extracting accurate 3-D spatial information has led to an unsustainable escalation of sensor hardware complexity/cost while producing diminishing returns. Perceptive is developing a set of novel techniques to extract and leverage additional layers of information - beyond simple spatial, 3D information - from sensing signals. In this talk we’ll discuss the state-of-the-art on these new ways to look at the world.