[ad_1]
Welcome to Episode 71 of The Robotic Report Podcast, which brings conversations with robotics innovators straight to you. Be a part of us every week for discussions with main roboticists, modern robotics corporations, and different key members of the robotics neighborhood.
In immediately’s episode, Steve and Mike interview Rashed Haq, VP of robotics for Cruise. Rashed discusses the present deployment of Cruise autonomous robotaxis in San Francisco and what the corporate expects to be taught from the expertise because it regularly expands its working scope.
He additionally discusses the varied challenges in creating and optimizing viable machine studying fashions that make choices within the protected operation of autonomous autos. We discuss all the use instances that should be thought of for a AV whereas it’s driving, and a number of the ways in which the Cruise robotics staff is dealing with them.
As an AI skilled and writer, we additionally discuss to Rashed concerning the distinctive nature of gathering coaching information and coaching new fashions to make use of in machine studying and AI-based algorithms. Rashed shares some priceless perception into the very best practices outlined in his e book and put into observe at Cruise.
Lastly, we discuss to Rashed about his visible artwork, together with the fascinating software of synthetic intelligence within the technology of computational images.
Steve and Mike additionally discuss concerning the newest robotics information tales from the final week.
Hyperlinks from the present this week:
Right here’s a pattern of certainly one of Rashed’s computation photography-based, AI-generated photos:
One picture from a sequence of laptop generated photos created by coaching an AI GAN after which asking it to generate a brand new picture. | Credit score: Rashed Haq
Together with his artwork, Rashed is exploring the rising space of computation images. Computational images leverages computer systems to routinely improve, enrich or generate photographic photos. Rashed leverages his ability as a machine studying engineer to create a brand new mannequin and practice a generative adversarial community (GAN) primarily based on his prior set of photos. With the GAN, Rashed can then create an algorithm that makes use of the data of the GAN to generate a model new picture from the set of photos that had been used to coach the GAN.
If you need to be a visitor on an upcoming episode of the podcast, or when you have suggestions for future friends or section concepts, contact Steve Crowe or Mike Oitzman.
For sponsorship alternatives of The Robotic Report Podcast, contact Courtney Nagle for extra data.
We wish your suggestions:
Inform us what you suppose are probably the most attention-grabbing robotics developments that may impression us in 2022? Depart us a voicemail.
[ad_2]