D2: Technologies, devices, circuits, and algorithms for neuromorphic event-based vision system
Half day, 13:30
The concept of neuromorphic events can be broadened beyond the conventional intensity change detections. For higher-level cognitive tasks, events of varying complexity are refined from simpler ones. The workshop introduces complex-event neuromorphic algorithms co-developed with sensors and computing hardware. Higher-level cognition algorithms operate with lower temporal rates than those designed for reflexes requiring immediate response to sensory data. Data transmission bottlenecks affect where simple events are transformed into more complex ones. Beyond-CMOS devices are widening the capabilities of the system in comparison to CMOS-only solutions. The workshop presents ideas and outcomes from the MISEL EU-project on neuromorphic sensing and computing, with additional insights from acknowledged experts in the field.
Ari Paasio (Kovilta Oy, FI)
Jacek Flak (VTT Technical Research Centre of Finland, FI)
Dr. Ari Paasio got this D.Sc. (Tech) degree in microelectronics from Helsinki University of Technology in 1999. Since then he has been working on integrated circuit design of massively parallel processors, thermal sensing readout optimization and ultra low-power IoT processors for e.g. medical sensors. He acted as a professor of Microelectronics at University of Turku, Finland, between 2003 and 2020. Since 2021 Paasio has worked full time at Kovilta Oy, a Finnish sensor-processor design company, where Paasio is one of the founders and a partner.