Rashed Haq on using AI to control autonomous vehicles and generate visual art

Welcome to Episode 71  of The Robot Report Podcast, which brings conversations with robotics innovators straight to you. Join us each week for discussions with leading roboticists, innovative robotics companies, and other key members of the robotics community.

In today’s episode, Steve and Mike interview Rashed Haq, VP of Robotics for Cruise. Rashed discusses the current deployment of Cruise autonomous taxi’s in San Francisco and what the company expects to learn from the experience as it gradually expanding its operating scope.

He also discusses the various challenges in creating and optimizing viable machine learning models that make decisions in the safe operation of autonomous vehicles. We talk about all of the use cases that have to be considered for a AV while it’s driving, and some of the ways that the Cruise robotics team is handling them. 

As an AI expert and author, we also talk to Rashed about the unique nature of gathering training data and training new models to use in machine learning and AI-based algorithms. Rashed shares some valuable insight into the best practices outlined in his book and put into practice at Cruise.

Finally, we talk to Rashed about his visual art, including the fascinating application of artificial intelligence in the generation of computational photography.

Steve and Mike also talk about the latest robotics news stories from the last week.

Links from the show this week:


Here’s a sample of one of Rashed Haq’s computation photography based, AI-generated images:

computer generated portrait by Rashed Haq

One image from a series of computer generated images created by training an AI GAN and then asking it to generate a new image. | Credit: Rashed Haq

With his art, Rashed is exploring the emerging area of computation photography. Computational photography leverages computers to automatically enhance, enrich or generate photographic images. Rashed leverages his skill as a machine learning engineer to create a new model and train a generative adversarial network (GAN) based on his prior set of images. With the GAN, Rashed can then create an algorithm that uses the knowledge of the GAN to generate a brand new image from the set of images that were used to train the GAN.

If you would like to be a guest on an upcoming episode of the podcast, or if you have recommendations for future guests or segment ideas, contact Steve Crowe or Mike Oitzman.

For sponsorship opportunities of The Robot Report Podcast, contact Courtney Nagle for more information.


We want your feedback:

Tell us what you think are the most interesting robotics trends that will impact us in 2022? Leave us a voicemail.

Credit: Source link

Comments are closed.