Students From HUGs Present Socially-Responsible AI Research At The exploreCSR Poster Session


Click the links that follow for more news about diversity and inclusion at Brown CS, exploreCSRour Socially Responsible Computing programAmy GreenwaldJeff HuangDaniel RitchieJames Tompkin, or other recent accomplishments by our students and faculty.

Thanks to an exploreCSR award from Google, Professors Amy GreenwaldJeff HuangDaniel Ritchie (lead organizer of the project), and James Tompkin of Brown CS have spent the past semester deploying a program for college students from HUGs in CS that exposed them to socially-responsible ways that AI can be used to realize creative visions. Working virtually, students from around the country became research associates, paired with Brown graduate student mentors from similar HUGs where possible to conduct individualized research experiences. (To read the news story that accompanied the program's launch, see here.)

"Many of these students didn't have any past exposure to research," says Amy. "And immediately they were participating in weekly research meetings – listening, asking questions, even presenting. Working one-on-one with a PhD mentor was invaluable for them."  

Last month, at 4 PM on April 30, the students met for the final time in a virtual poster session where they presented their work. 

Hana Memon, a CS honors student at The College of New Jersey, worked with James to explore privacy using StyleGAN, a technology often used to spread misinformation using deepfakes. In an interesting reversal, their research team took a new approach, using it to protect one's digital identity. By using StyleGAN alongside Python and OpenCV, they created a system that was able to take in a photo, recognize a face, and produce a new image with slightly altered facial features.  

"While we only touched the surface of what this new technology can do," she says. "I’m grateful to have been exposed to the vast world of computer science research. As a first-year CS student, I found this program particularly powerful because I was able to combine my creative and computational interests for the first time. Not only was I able to learn about computer vision and HCI directly from the experts in the field, but I was also guided on how to write about and present my work through the various workshops I attended. After seeing my peers’ work at the final symposium, I’m even more excited to use my newfound skills to use AI to reimagine computational creativity."

Sadie Zhao, a student from Pomona College, says that the program had a profound impact on her academic development. "Before working [with Amy]," she says, "I was interested in theoretical topics but struggled to start. After attending the reading group, I learned how to study through the process of reading, listening, and discussing. [Amy] showed me a whole new path to improving myself and ignited my excitement about the subject."

"For my research project," explains Kelly Ly of University of Massachusetts, Lowell, "I explored texture reconstruction from 2D images to 3D objects, identified what gaps in texture generation models exist in research, and learned how to address those gaps. I learned so much about 3D modeling, deep learning, open source code, and graduate school—all things I didn’t know much about until attending this program. I had the opportunity to meet students from all across the nation and I loved hearing about the amazing work they’ve accomplished at the end-of-semester research symposium."

Tongyu Zhou was one of the PhD mentors. "It was super fulfilling introducing a younger student to research," she says. "My mentee had always wanted to enter academia in the future, but never really knew how to start. We were able to take an initial stab at a difficult problem; she came up with creative ways to solve problems via means I hadn't previously thought of, and was always eager to consider the implications of solutions. I'm excited to see where she goes next."

It's an excitement shared by the other mentors and faculty members, who see the program as a behind-the-scenes look at how CS research operates, normalized and without pretentions, that students can now use as a building block for future careers. Due to the COVID-19 pandemic, this year's program was entirely virtual, but Amy, Daniel, James, and Jeff are considering possible alternatives for future years that might include a hybrid program with an in-person event at the end of the semester.

"It was wonderful to get an excited Slack message from one of my mentees the first time she got some interesting results," Daniel says. "I was amazed by how dedicated the students were, and how much work they wanted to put in. This was a really 'you choose how deep you want to go' kind of program, and all of my mentees did a massive amount of work over the semester." 

For more information, click the link that follows to contact Brown CS Communication Outreach Specialist Jesse C. Polhemus.