News 2018

July 2018

Hodgins, Gupta Join Facebook AI Research

Byron Spice

Carnegie Mellon University's Jessica Hodgins will lead a newly established Facebook AI Research (FAIR) lab in Pittsburgh, where she will be joined by CMU's Abhinav Gupta. The appointments are part of an expansion of Facebook's artificial intelligence research activities with academic communities. The FAIR Pittsburgh lab will focus on robotics, lifelong-learning systems that learn continuously, teaching machines to reason and AI in support of creativity. Yann LeCun, Facebook's chief AI scientist, announced the Pittsburgh lab and the expansion of FAIR labs in Seattle, London and Menlo Park, Calif., in a blog post today. Hodgins, a professor of robotics and computer science, and Gupta, an associate professor of robotics, will work part-time with Facebook, maintaining their teaching and research duties at CMU. This dual-affiliation model is common across FAIR, LeCun noted, with researchers already splitting their time between FAIR and New York University, Georgia Tech, the University of Montreal, the University of Washington, University College London and Tel Aviv University, among others. "Facebook's new lab will create new opportunities to make advances in AI, both at the company and at CMU," said Andrew Moore, dean of the School of Computer Science. "Even better, both Jessica and Abhinav will continue their work at CMU, helping us educate the next generation of AI scientists." Similarly, Yaser Sheikh, CMU associate professor of robotics, has divided his time since 2015 between campus and Facebook, where he is now director of the Facebook Reality Lab. LeCun said FAIR plans to support a number of Ph.D. students and provide research funding for Carnegie Mellon students and faculty. In addition to collaborations, FAIR also offers fellowships and emerging scholars programs for doctoral students. "We created Facebook AI Research over four years ago to focus on advancing the science and technology of AI, and we've always done this by collaborating with local academic communities," LeCun said in his blog. "FAIR relies on open partnerships to help drive AI forward, where researchers have the freedom to control their own agenda. Ours frequently collaborate with academics from other institutions, and we often provide financial and hardware resources to specific universities." Carnegie Mellon is a leading center for AI research and education, and was ranked number one for AI graduate programs by U.S. News and World Report earlier this year. The CMU AI initiative is further strengthening the university's AI efforts by connecting and supporting AI researchers across campus. This fall, the School of Computer Science will launch an undergraduate AI degree — the first at any U.S. university. Hodgins, formerly vice president for research at Disney Research, focuses her work on computer graphics, animation and robotics with an emphasis on generating and analyzing human motion, which will also benefit the Facebook Reality Lab. Hodgins last year was elected president of the Association for Computing Machinery's Special Interest Group on Computer Graphics and Interactive Techniques (SIGGRAPH). Last summer, SIGGRAPH presented her with its highest honor, the Steven Anson Coons Award for Outstanding Creative Contributions to Computer Graphics in recognition of her foundational work in character animation, her support and cultivation of emerging researchers, and her extensive volunteer service to the computer graphics community. Gupta's research focuses on large-scale visual and robot learning, self-supervised learning, and reasoning. A faculty member since 2011, he has received the Office of Naval Research's Young Investigator Award, a Sloan Research Fellowship, a Bosch Young Faculty Fellowship and the Okawa Foundation Research Grant.

Architect Reveals Design of TCS Hall

SCS Will Be an Occupant When Building Opens in 2020

Byron Spice

The architectural firm Bohlin Cywinski Jackson (BCJ) has released the design for TCS Hall, which will house computer science offices and labs when it opens in spring 2020. Faculty and staff from both the School of Computer Science and the Department of Mechanical Engineering will occupy the 90,000-square-foot building. It is funded in part by a gift from Tata Consultancy Services, a global IT and business solutions company. Constructed on a site between 4615 Forbes Avenue and Junction Hollow, TCS Hall will serve as a gateway to campus for those entering from Forbes and as a keystone for further development north of Forbes. The building has been redesigned to almost double its size since ground was initially broken in 2017. BCJ says the building will feature a "Student Collaboration Porch," a drone highbay, flexible makerspace and an outdoor robot yard. Both TCS Hall and ANSYS Hall, now under construction next to Hamerschlag Hall, are projects led by BCJ's Gregory Mottola, design principal, and Kent Suhrbier, principal in charge. Mottola and Suhrbier are alumni of CMU's School of Architecture. Other works by BCJ for CMU include the Stever House, the Software Engineering Institute and the Robert L. Preger Intelligent Workplace. The firm's local projects include the Heinz History Center and the Frick Environmental Center.

SCS Faculty Receive NSF CAREER Awards

Byron Spice

Bernhard Haeupler, Louis-Philippe Morency and Jean Yang are the latest School of Computer Science faculty members to receive the National Science Foundation’s prestigious Faculty Early Career Development (CAREER) award. They are among 307 CAREER recipients this year in computer science and engineering in which the NSF has invested $150 million over the next five years, placing their academic careers on firm scientific footing and giving them the opportunity to serve as academic role models in research and education. Those recipients include Claire Le Goues of the Institute for Software Research and Brandon Lucia of the Computer Science Department (CSD) and the Electrical and Computer Engineering Department, whose awards were announced earlier this year. Haeupler, an assistant professor in CSD, will receive $560,000 to develop a theory and methods for correcting errors in interactive communications. Coding theory for reliable one-way data transmissions over unreliable channels is well-established, but development of error-correcting codes for the interactive communications that typify modern systems is a much harder task. Haeupler joined CSD in 2014, after a one-year stint as a post-doctoral researcher at Microsoft Research Silicon Valley. He earned his Ph.D. in computer science at MIT, where his dissertation won the George Sprowls Award for best computer science Ph.D. thesis and, later, the 2014 ACM-EATCS Doctoral Dissertation Award for Distributed Computing. Morency, assistant professor in the Language Technologies Institute and director of the Multimodal Communication and Machine Learning lab, will receive $550,000 to develop technologies that allow computers to understand subtle nonverbal behaviors of people and to learn the inherent variability between individuals in how those behaviors are expressed. These technologies could aid doctors in assessing mental disorders, help animators develop characters whose behaviors match their personality and assist educators in measuring student engagement. Morency, who earned his Ph.D. in computer science at MIT, joined LTI in 2015 after serving on the research faculty of the University of Southern California. Earlier this year, he received a three-year Finmeccanica Career Development Professorship in Computer Science. IEEE Intelligent Systems named him one of ''AI’s 10 to Watch'' in 2008. He also received the NetExplo 2014 Award, presented in partnership with the U.N. Educational, Scientific and Cultural Organization (UNESCO), as one of the year’s 10 most promising digital initiatives. Yang, an assistant professor in CSD and affiliated faculty in the Computational Biology Department, was awarded $550,000 to develop a new programming model that incorporates a theory of differential privacy. Differential privacy protects individual data values while allowing the release of results of analyses of the data. She proposes a programming framework called Jostle that makes it easier for programmers to implement privacy-preserving analytics by allowing the compiler and runtime, rather than the programmer, to make algorithmic choices that satisfy privacy requirements. Yang earned her Ph.D. in computer science at MIT in 2015; she joined CSD in 2016 after serving as a post-doctoral researcher at Harvard Medical School. MIT Technology Review named her to its 2016 list of Innovators Under 35 in recognition of her work on programming models that incorporate security. ''We’re delighted to support this cadre of early-career researchers as they embark on long-term research and education activities that will advance the frontiers of our field,'' said Erwin Gianchandani, acting assistant director for Computer and Information Science and Engineering (CISE) at NSF. ''These early-career faculty will catalyze new breakthroughs in computer and information science and engineering that will transform our nation in the years to come.''