Welcome to Talking Autonomy – Ghost’s new tech talk video series! Talking Autonomy features the builders and do-ers behind autonomy – the founders, engineers, designers, mathematicians, and even the policy-makers who are responsible for bringing Ghost’s self-driving technology to the roads. The series aims to provide unscripted and unabashedly technical insights into both Ghost’s core technologies and the autonomy market as a whole.
In this inaugural episode we’re joined by Ghost co-founder and CTO Volkmar Uhlig, exploring how KineticFlow, Ghost’s vision-based perception neural network, was inspired by the biological systems that allow the human eye and brain to sense distance, motion, and velocity. Volkmar discusses the various human pathways for visual perception, and how some of these can be implemented into neural networks to enable AI-based computer vision for autonomy.
And this video is, of course, just the beginning of Talking Autonomy. Tune in here roughly weekly for new topics and insights, including next week’s episode where we’ll explore the “long tail” of image recognition, and why building AI that isn’t tricked by uncommon actors and difficult scenarios is so challenging.