Data Science Meetup - Apr2020 - Complexities of Computer Vision
*This event was hosted by our Data Science, Machine Learning & AI Meetup Group: https://www.meetup.com/Charlotte-Data... *
SPEAKER
Richard Hurst - Director of AI at Camgian Microsystems
TOPIC
Complexities of Computer Vision within Uncertain Operational & Physical Environment
Within the Computer Vision domain, MPRI (Machine Perception, Reasoning & Intelligence) enhances the capabilities of autonomy by providing technologies and needed information for a system to better understand the mission environment and how to adapt, in-situ, when the predicted environment differs from the a priori data provided. This is ever so evident when searching for mine-like objects in a very cluttered and murky subsurface environment. Through various computer vision, data procurement and fusing techniques, reasoning, and knowledge-based systems can be leveraged to assess miscellaneous and potentially harmful objects on the seabed floor, unanticipated scenarios, make recommendations (or even real-time decisions in a contested or communication denied environment), and over time, learn from each experience in such a way that improves the probability of success in future unanticipated encounters. In short, enabling an autonomous system to adapt and react within an environment, performing situational dependent tasking, in a near human-like manner. This discussion will cover the basics of Computer Vision and why it’s not as easy to achieve ideal results as one might think.