Keynote talk: Embedded vision systems

Summary form only given, as follows. While machine vision research has continued to improve multi-fold over the past few decades, it still significantly falls short of the abilities and efficiencies of the primate visual cortex system. The primate brain is especially superior as pertains to comprehending and interacting with complex natural environments. In energy-efficiencies, the brain is estimated to be five to six orders better than current machine vision solutions. While there is much consensus on the superiority of biological vision systems over machine systems on most vision tasks, the approaches leading to the better efficiencies and flexibility akin to the visual cortex are still widely debated. In this talk, I will highlight recent efforts at architecting customized digital hardware systems using neuromorphic algorithms as one successful approach to achieving better energy efficiencies. I will also outline emerging embedded visions applications.