Amazon had a vision. Gerard Medioni had a computer vision. Together, there’s apparently no stopping the tech giant and the computer scientist who serves as director of research for Amazon Go, the company’s off-and-running cashier-less store idea.
Medioni, who is our latest Geek of the Week, has been a professor at the University of Southern California for 35 years, and previously chaired USC’s Computer Science department. While he still teaches at the school, his world (and location) changed when he joined Amazon in 2014 to work on Amazon Go.
“In collaboration with an incredible team, I helped turn the concept into reality, with a functioning store open to the public since January 2018, and more on the way,” Medioni said.
In fact, Amazon has four of the stores in Seattle now, three in Chicago and one in San Francisco. The latest in Seattle, located on the sixth floor of the downtown Macy’s building, is a compact version, revealed earlier this week.
Medioni has worked with several companies in the past, including Symah Vision where he developed the first system to insert virtual ads into live video for sports events (the 10-yard yellow line during football games is an example). He also worked at Prime Sense for seven years, where he helped develop the sensor used in the Microsoft Kinect, which he noted “still holds the title of fastest-selling consumer device, according to Guinness World Records.”
When he’s not working with his Amazon team on advancing cutting edge shopping experiences, Medioni also organizes and chairs computer vision conferences — CVPR and ICCV in particular — and edits a book series, “Synthesis Lectures on Computer Vision” for Morgan & Claypool Publishers.
Learn more about this week’s Geek of the Week, Gerard Medioni:
What do you do, and why do you do it? My role is to invent and create technology that works so seamlessly, it appears to be magic. One of the challenges is to tell the difference between what is “very difficult” from what is “impossible.” The work I do with Amazon Go’s Just Walk Out Technology lies at what I call “the edge of the doable.” It’s been the proudest thing I have worked on in my career to date.
What’s the single most important thing people should know about your field? Vision is our most powerful (and most complex) sense, and Computer Vision aims to emulate what our eyes and brain together perform so seamlessly and naturally. The goal of Computer Vision is to transform an array of colored dots into semantic information, such as objects and people, as well as allow us to perceive the environment and interact with it. Applications of working Computer Vision systems are both broad and impactful, such as self-driving cars and flying drones, smart robots, medical diagnostic aids, visual assistance for the blind and more.
This said, in spite of the extraordinary recent progress, computer vision is far from being “solved.” As we often say at Amazon, we’re just at “day one” of this technology. A common misconception I often face is that people think computer vision should be easy, simply because people do it with no apparent effort. It’s much more complex in actuality. As we continue to push the state of the art forward in computer vision and deep learning, we are growing our team of researchers to work on many interesting challenges in these spaces like how to continue to train our algorithms to use less data to make decisions, how to apply learning from one area to another and beyond.
Where do you find your inspiration? I find inspiration in science fiction. Authors such as Jules Verne, Isaac Asimov, Philip K. Dick imagine worlds of the future — it’s my perspective that it’s our role to try to fulfill the visions (at least the non-apocalyptic ones!).
What’s the one piece of technology you couldn’t live without, and why? I have come to increasingly depend on my cell phone, as it has earned the “smart” phone label. Besides being a telephone, it allows me to access information on demand, make travel reservations, purchase goods, guide me on the road, provide entertainment in the form of games, music and video as well as get me into the Amazon Go store! This vastly exceeded the expectations I had made 20 years ago about it.
What’s your workspace like, and why does it work for you? My desk is on the messy side. It has several piles of papers and books, reflecting my tendency to multi-task and keep multiple threads open simultaneously. At Amazon, we review ideas and plans in narrative form, so I always have a lot of plans around me at my desk.
Your best tip or trick for managing everyday work and life. (Help us out, we need it.) The best tip is to enjoy what you do at work, which makes you look forward to coming to the office. Second, is to maintain a rich social life, especially with close family and friends. Finally, find some way to exercise that makes you want to do it again; for me it is playing soccer.
Mac, Windows or Linux? At work, Windows, mostly because I need to play my old presentations with embedded movies, which never really worked on the Mac. At home, I use Mac because it is so much more intuitive for naïve users (my family…).
Kirk, Picard, or Janeway? Picard. Use logic and data to get to a decision, delegate whenever possible, and negotiate rather than fight. And he is French, like me.
Transporter, Time Machine or Cloak of Invisibility? Time machine but as passive observer, as changing the course of history gets horribly complicated very quickly. I’d use it to find out what and how events really happened, instead of what we read in books. I’d also like to zip through time to witness geological scale changes.
If someone gave me $1 million to launch a startup, I would … Use it to define an application which would be a game changer for its target audience, build a prototype, get feedback from that population, refine the concept, then pitch it to raise real money and build a company around it. For instance, I would propose to fully develop the prototype I built at USC to enable blind people to navigate both indoors and outdoors in unknown environments.
I once waited in line for … Tickets to the rock band The Who, on their Los Angeles tour appearance at the Sports Arena. I got there hours before the box office doors opened, and did get the tickets, which sold out in minutes.
Your role models: The people I admire most are:
- Albert Einstein, whose intellectual curiosity made him question the very essence of what his teachers were presenting as dogma and explained our universe using thought only, years or even decades before his theories were experimentally verified.
- Steven B. Sample, who served as USC president and transformed the school into an elite institution through his vision, focus, and execution excellence. His book, “The Contrarian’s Guide to Leadership” continues to serve as a useful guide to me.
- Finally, I tip my hat to Jeff Bezos, who is one of the very few leaders today with a vision beyond the immediate future. He has created not only a company, but also a culture and a way of thinking.
Greatest game in history Bridge. It is a fascinating game that combines a variety of skills both in the bidding and card playing stages, with team work and trust in your partner. The duplicate form of bridge, which I prefer, essentially eliminates chance and the quality of the cards you are dealt.
Best gadget ever: The film camera, which was followed by the digital camera, which is now replaced by the ubiquitous smartphone to take pictures.
First computer: Vax 750. Am I dating myself? Do I miss teletype and punched cards, with the dreaded “fatal error” after a nightly run? Not really.
Current phone: iPhone 7. I should upgrade since the battery is starting to fail.
Favorite app: Waze.
Favorite cause: I am a big fan of the Chabad organization. They provide help to a large number of people throughout the world, regardless of political or religious affiliations. For instance, in Los Angeles, the Chabad Residential Treatment Center is an inpatient, residential, drug and alcohol addiction and dependence treatment center, and detox facility. I have witnessed firsthand the impact they have in bringing people from the abyss, to a productive and positive life.
Most important technology of 2018: Amazon Go’s Just Walk Out Technology, of course! Kidding aside, our Just Walk Out Technology is truly something I’m very proud to be a part of. Taking the vision of having customers take what they want from the store and just walk out was an incredible challenge and to see the store providing customers with such an effortless shopping experience is so rewarding to see.:
Most important technology of 2020: A tossup between spaceflight services and self-driving automobiles.
Final words of advice for your fellow geeks: Don’t be afraid to think big, and then work hard to get there, but remember to have fun along the way.
Website: Amazon Go
LinkedIn: Gerard Medioni