News

Algorithm probes how AIs reason

Quartz  explores an algorithm devised by CS Prof. Trevor Darrell, L&S CS undergraduate student Dong Huk Park, CS grad student Lisa Anne Hendricks, and postdoc Marcus Rohrbach, along with researchers in the Max Planck Institute for Informatics,  in an article titled "We don’t understand how AI make most decisions, so now algorithms are explaining themselves." Engineers have developed deep learning systems that ‘work’ without necessarily knowing why they work or being able to show the logic behind a system’s decision.   The algorithm uses a “pointing and justification” system, to point to the data used to make a decision and justify why it was used that way.

Computational Imaging proposal accepted for collaborative research initiative

A Computational Imaging research proposal submitted by EE Associate Prof. Laura Waller, EE Associate Prof. Michael Lustig, CS Assistant Prof. Ren Ng, CS Assistant Prof. Jonathan Ragan-Kelley, and CS Associate Prof. Benjamin Rechts has been accepted as part of a set of cross-disciplinary activities planned for development by Berkeley Research.  Berkeley Research ran eight faculty forums on a wide range of topics and received 30 proposals which were reviewed by a faculty panel and discussed with the Deans.  The selected projects "hold great promise for Berkeley to be at the forefront of developing a positive vision for the future."

Center for Advancing Women in Technology logo

Center for Advancing Women in Technology launches Technology Pathways Initiative

Center for Advancing Women in Technology (CAWIT) in collaboration with  U.C. Berkeley, San Francisco State University and San José State University, through $3M in investment from Intel Corporation, KLA-Tencor Foundation, and Salesforce, will launch the Technology Pathways Initiative (TPI), to increase participation of women in CS fields through the development of new interdisciplinary CS degree programs at three pilot campuses in 2017. Prof. Tsu-Jae King Liu has been developing the Women In Technology workshop at UC Berkeley.

Berkeley AI Research Lab logo

NVIDIA Delivers AI Supercomputer to Berkeley

Earlier this year NVIDIA CEO Jen-Hsun Huang delivered a NVIDIA DGX-1 AI supercomputer in a box to the Berkeley AI Research Lab (BAIR). BAIR’s research is at the cutting edge of multi-modal deep learning, human-compatible AI and connecting AI with other scientific disciplines and the humanities. According to Prof. Pieter Abbeel, “More compute power directly translates into more ideas being investigated, tried out, tuned to actually get them to work.”

RISC-V (Five) is Alive!

RISC-V, an open-source instruction set architecture created at UC Berkeley is featured in an electronic design article titled “RICS-V (Five) is Alive!” RISC (Reduced Instruction Set Computer) was originally designed in 1982 by students with the direction of Professors David Patterson and Carlo Sequin. Since then, iterations of RISC have been developed. In 2010 Prof. Krste Asanovic, with the help of Prof. Patterson, decided to develop another version of RISC to help both academic and industrial users and RISC-V was published.

Ken Goldberg named Chair of IEOR

Effective January 1, 2017, Prof. Ken Goldberg will serve as Chair of the Department of Industrial Engineering & Operations Research (IEOR).  Ken is a professor of IEOR, with secondary appointments in EECS, Art Practice, the School of Information and Radiation Oncology at UCSF’s Medical School. Ken is Director of CITRIS’s People and Robots Initiative and UC Berkeley’s AUTOLAB, where he and his students pursue research in geometric algorithms and machine learning for robotics and automation in surgery, manufacturing and service applications.

Ruzena Bajcsy and Mike Stonebraker are among the "7 over 70"

A companion to Tech Review’s annual 35 Innovators Under 35 list features a list of seven innovators over 70. The new list includes EECS Professor Ruzena Bajcsy and professor emeritus Michael Stonebraker.  The 7 Over 70 list acknowledges innovators who are continuing to have sustained impacts in their field well after most of their colleagues have decided to retire.

prof. david wagner

David Wagner receives ACM SIGSAC 2016 Outstanding Innovation Award

Prof. David Wagner has won the ACM Special Interest Group on Security, Audit and Control (SIGSAC) 2016 Outstanding Innovation Award. This award is given for outstanding and innovative technical contributions to the field of computer and communication security that have had lasting impact in furthering or understanding the theory or development of secure systems. Prof. Wagner is recognized "For innovative research in systems security, software security, and cryptography that has inspired research in sandboxing, static analysis for security, and cryptanalysis."

amplab and riselab logos

AMPLab ends, RISELab begins

After six years of delivering major technological advances like Apache Spark, Apache Mesos and Alluxio, the AMPLab (Algorithms, Machines and People Lab) directed by Adjunct Prof. Michael Franklin and Profs. Michael Jordan and Ion Stoica will be closing and the RISELab (Real-time Intelligent Secure Execution Lab) will take it’s place. The RISELab will continue the work of the AMPLab, tackleing the next phase in distributed computing. Prof. Ion Stoica will continue his role as director and will be joined by Prof. Joe Hellerstein and Assistant Profs. Joseph Gonzalez and Raluca Ada Popa. AMPLab End of Projects events will be held on Nov. 17 & 18, 2016 at the International House, UC Berkeley.

Meet the professor who will help robots learn common sense: Sergey Levine

Computer Science Assistant Prof. Sergey Levine is the subject of an article in BGR about machine learning titled Meet the professor who will help robots learn common sense.  “One of the things I think we’ve seen with computer vision is the bottom-up approach tends to be very effective,” Levine says. “In other words, once you figure out a good way to acquire the low-level representations — in the case of vision, things like the fact that images consist of edges — then whatever technique you use that’s general that can acquire those low-level representations will also be able to deal with the higher level stuff”

“So for me, part of the hope is if we can find the right way to acquire the low-level behaviors, the higher behaviors will begin to emerge naturally. Using the same technique just applied at a larger scale.”