A team led by EECS Associate Professor Boubacar Kanté and EECS Professor Eli Yablonovitch has developed a machine-learning based optimization method for nonlinear and quantum optics. Inverse-design has been traditionally applied to linear optical systems, and it often leads to optimized structures that are unintuitive or experimentally unrealistic. In this study, published in Optica, the researchers attempt to tackle these challenges using a new inverse-design method for nonlinear photon generation. According to lead author and graduate student Zhetao Jia, they were able to achieve a compact, robust, and efficient source of entangled photon pairs based on spontaneous four-wave mixing in silicon, the most common material used in the semiconductor industry. This nonlinear quantum-optics approach could potentially be used for large-scale communication and quantum computing applications.

CS Associate Professor Alvin Cheung has won the 2023 Very Large Data Bases (VLDB) Early Career Research Contribution Award. The award, which includes a $2,000 prize, recognizes researchers who have made a significant impact through a specific contribution to the field since completing their Ph.D. Separately, a paper by CS Professors Joseph Gonzalez and Joseph Hellerstein, co-authored by Yucheng Low, Aapo Kyrola, Danny Bickson, and Carlos Guestrin, received the 2023 VLDB Test of Time Award. Their paper, "Distributed Graphlab: A framework for machine learning in the cloud," was published at VLDB 2012. The authors were nominated for this award by the research community, and the winner was selected based on the paper's impact through its consequent products and services, and follow-through research by the community. The VLDB awards recognize excellence in the field of database research and development. The awards are presented annually at the VLDB conference, which is one of the premier conferences in the database field.

EECS graduate students Yicheng Zhu and Jiarui Zou, and post-docs Ting Ge and Nathan Ellis have won the 2023 IEEE Control and Modeling for Power Electronics (COMPEL) Best Paper Award. Their paper, "A 48-V-to-1-V Switching Bus Converter for Ultra-High-Current Applications,” demonstrated a new dc-dc power converter topology and control technique for data center power delivery applications, capable of sourcing 1200 A of current at 1 V supply voltage. The hardware prototype used to validate the concept achieved the highest power density and efficiency combination of any prior work, academic or industrial. Next, the researchers are working with industry partners to transition this record-breaking concept to next-generation GPU/CPU computing platforms for AI and machine learning applications. IEEE COMPEL is the premier control and modeling conference for power electronics, having brought together world experts in the field for the last 24 years. Three best papers were selected this year from the total accepted 84 papers, based on originality, contribution to the field, and quality of presentation at the conference. 

A team of researchers from UCSF and Berkeley EECS have developed an implantable AI-powered device that can translate brain signals into modulated speech and facial expressions. The device, a multimodal speech prosthesis, and digital avatar, was developed to help a woman who had lost the ability to speak due to a stroke. The results have the potential to help countless others who are unable to speak due to paralysis or disease. The breakthrough study, published in the journal Nature, was led by UCSF neurosurgeon Edward Chang, EE Assistant Professor Gopala Anumanchipalli and Ph.D. student Kaylo Littlejohn. “This study heavily uses tools that we developed here at Berkeley, which in turn are inspired by the neuroscientific insights from UCSF,” said Gopala. “This is why Kaylo is such a key liaison between the engineering and the science and the medicine — he’s both involved in developing these tools and also deploying them in a clinical setting. I could not see this happening anywhere else but somewhere that is the best in engineering and the best in medicine, on the bleeding edge of research.”

A team led by CS Assistant Professor Anjoo Kanazawa has created Nerfstudio, an open-source platform to help speed up the development of Neural Radiance Fields (NeRFs). NeRFs are a type of 3D imaging technology that can be used to create photorealistic 3D models of objects and scenes from a series of images. The plug-and-play framework, called Nerfstudio, makes it easier for researchers to create and train NeRFs, allowing users to run NeRFs on real-world data. “Advancements in NeRF have contributed to its growing popularity and use in applications such as computer vision, robotics, visual effects and gaming. But support for development has been lagging,” said Kanazawa. “The Nerfstudio framework is intended to simplify the development of custom NeRF methods, the processing of real-world data and interacting with reconstructions.”

EE Professor Venkat Anantharam and Chinese University of Hong Kong (CUHK) Professor Cheuk-Ting Li have won the 2023 IEEE Information Theory Society Paper Award. The award is given annually “for an outstanding publication in the fields of interest to the Society appearing anywhere during the preceding four calendar years.” The paper “A unified framework for one-shot achievability via the Poisson matching lemma,” by Li and Anantharam, appeared in the IEEE Transactions on Information Theory in February 2021, when Li was a postdoctoral fellow at Berkeley EECS. Li is now an assistant professor at the CUHK. Anantharam, who is an IEEE Fellow, received this award once before in 2008 for the paper “Bits Through Queues.”

An article produced by the College of Computing, Data Science, and Society highlighted the increasingly frought landscape of user privacy in the emerging world of Virtual Reality (VR) devices. The article cites two papers published by faculty, students, and visitors affiliated with the Berkeley Center for Responsible, Decentralized Intelligence. Led by CS Ph.D student Vivek Nair and Professor Dawn Song, the research showed that users of such devices can be identified using just minutes of their head and hand movements. Movement data, which is collected and shared with companies and other players to fuel these worlds, can be used to infer dozens of details from age to disability status. One paper demonstrates that body movements are as singular and reliable an identifier as fingerprints, which was accepted for publication at the USENIX Security Symposium. Another found that use of headset data could accurately identify or infer more than 25 characteristics, including location, age and height, which will be published for the Privacy Enhancing Technologies Symposium. “We've done an extensive job of proving that there is a privacy risk here and that it is a different kind of privacy risk than what we have seen on the web,” Nair said. “These kinds of approaches for how to either transform the data or control who has access to it, that's going to be our main focus moving forward." Berkeley RDI is a multi-disciplinary initiative aimed at advancing the science, technology and education of decentralization and empowering a responsible digital economy. This work is part of the center’s Metaverse security and privacy research effort.

A team of researchers led by Professor Boubacar Kanté has demonstrated the first on-demand quantum light source using silicon, an advancement towards creating photons in ways that would reliably feed quantum networks, or a quantum internet. “The possibility to use silicon as a source of quantum light signifies that current large-scale Complementary Metal-Oxide Semiconductor (CMOS) chip manufacturing processes at the core of today’s optoelectronics and artificial intelligence (AI) devices may be directly used for future quantum systems,” Kanté said. He elaborated further: "In this work, we successfully embedded for the first time an atomic defect in silicon the size of atoms (1 angstrom) in a silicon photonic cavity (1 micron) with the size of less than one-tenth of a human hair. The cavity forces the atom to be brighter, and it emits photons at a faster rate. Those are necessary ingredients for scalable quantum light sources for the future [quantum] internet." This research was published in Nature Communications on June 7th, 2023. The study was led by post-doctoral scholars Walid Redjem and Wayesh Qarony, and Yertay Zhiyenbayev, a third-year Ph.D. student in Kanté’s group. Other co-authors include Schenkel, Vsevolod Ivanov, Christos Papapanos, Wei Liu, Kaushalya Jhuria, Zakaria Al Balushi, Scott Dhuey, Adam Schwartzberg and Liang Tan. The National Science Foundation and the Department of Energy provided the primary support for the study. Additional funding came from the Office of Naval Research, the Moore Inventor Fellows program and UC Berkeley’s Bakar Fellowship.