Peter Zatloukal, Vice President of Engineering at Xnor.ai, demonstrates the company's latest embedded vision technologies and products at the 2019 Embedded Vision Summit. Specifically, Zatloukal showcases new AI technologies that reshape potential for the industry.
Zatloukal begins by demonstrating how Xnor's deep neural networks are capable of running on edge devices as small and low-cost as a Raspberry Pi Zero. Next he showcases AI2GO, a new self-serve platform to deploy thousands of Xnor's pre-trained models, tuned for a variety of use cases and performance characteristics across different hardware targets. Lastly he introduces Xnor’s Software-Defined Sensor, a novel FPGA-based solution co-designed with Xnor's deep learning models to deliver high performance AI at the lowest power ever seen.