WRITTEN BY: Matt Busekroos

Ethan Weber has always loved building projects and it’s been a favorite hobby of his for as long as he can remember. Weber is from New Holstein, a small, rural city in Wisconsin. While Weber’s school didn’t emphasize STEM, his parents supported his interests and purchased him electrical components in middle school, such as Arduinos and Raspberry PIs, that he could use to build contraptions, such as small robots.

In high school, Weber discovered the FIRST Robotics Competition (FRC), so he joined the nearest team and found a community of people similar to himself. Weber bonded with like-minded, motivated peers who he could collaborate with. Weber was drawn to MIT for the same reasons.

Weber is now a Master’s student working with Professor Antonio Torralba. His current research projects with Torralba’s lab include depth prediction with multi-view invariant constraints, efficient instance segmentation dataset creation and damage assessment in imagery. When Weber became interested in robotics perception, he realized how related the field of augmented reality is, which share many of the same research problems, but different applications. He is hoping to pursue a PhD starting next year after his MEng. Weber’s goal is to focus on augmented reality and start a company in the space after finishing his PhD.

“I'm excited by AR because of the seamless integration it will offer between the digital and physical world,” Weber said. “Currently we interact with 2D projections (phone screen, TV, computer, etc.), but by understanding the 3D world from multi-view images, we will be able to transition into the AR future where the digital and physical worlds can coexist.”

Weber said he plans to focus his PhD on this area and iteratively work on the major limitations in AR from a computer vision standpoint. He said he enjoys many aspects of computer vision.

He plans to participate in collaborations and contracting opportunities in impactful areas that involve the use of computer vision, or image understanding. He said AR is the application he’s most interested in because it's a tool that can be applied to many problems and it will lead to more creative, improved connectivity and better visualizations of data.

“I care a lot about the downstream applications and impact of my work,” Weber said.  “Furthermore, I make decisions based on the ‘delta’ that I can make. I'm motivated when I'm in a position to maximize my relative impact”.

Weber worked on an application to improve posture from laptop webcams and he has worked on natural disaster understanding with computer vision. He is interested in areas where he can expand beyond current expectations (or current results). He also enjoys working collaboratively to make a greater impact.

“I enjoy working when I can push ideas forward quickly and see how they can impact others. I also love the thrill of working quickly with a team,” Weber said.

Prior to working with Torralba, Weber worked on robotics with Professor Brian Williams for one year and Professor Russ Tedrake for two years. Working across robotics disciplines helped Weber determine which research area appeals most to him.

“I realized that of the parts of robotics that I enjoy, I'm most interested in perception and getting robots to see and understand the world,” Weber said. “This transition from robotics generally to robotics perception made Professor Antonio Torralba's lab very interesting to me, so I reached out and was incredibly fortunate to join my MEng.”

Weber said working with Professor Antonio Torralba has been an incredible experience.

“He frequently checks in, and he's helped me to work on three projects in the lab now,” Weber said. “I couldn't ask for a better experience with the lab and the insights I've gained from Professor Antonio Torralba, especially with his questions that help guide research projects.”