CS News summer 2024

SUMMER 2024

ACTION Smart Bionic Eye INSTITUTE

Human–AI Integration

AWARDS & HONORS 1

1,000+ cluster brings low-powered supercomputing to UCSB.

Smart Bionic Eye

Neuroscience meets AI with support of new NIH grant.

10,794 / 972

So Much Raspberry Pi

8000 600 6000 300 4000

0 2000 '13 '14 '15 '16 '17 '18 '19 '20 '21 '22 '23

0

'13 '14 '15 '16 '17 '18 '19 '20 '21 '22 '23

Freshmen Enrollment

Alferness-Henke

175

Donor David Henke endows new chair for UCSB Computer Science.

150 1200 125

Capstone 2024

100 900

Student teams find innovative solutions to industry problems.

Welcome Party

Get ready to get down with CS faculty and friends.

Broadening Participation

75

133

Teaming humans and AI to fight cyberattacks.

10000 900

600 50 25 300 0

'13 '14 '15 '16 '17 '18 '19 '20 '21 '21 '23

'13 '14 '15 '16 '17 '18 '19 '20 '21 '22 '23 Total Enrollment 0

Graduate

Representation, relevance, and equity in computer science.

800

Human-AI Integration

600175

Expanding opportunities for all with interactive AI + XR.

150 400 125

PhD Application Support

100 200 75

Pre-application review service aids underrepresented students.

Awards & Honors

Faculty and students recognized for their achievements.

Undergraduate

194

4 5 6 8 10 12 14 16 18 20

ACTION Institute

1200

Freshmen Applications/Admits

663

Contents

CS@UCSB Enrollment

0 50 '15 25

'16

'17

'18

'19

'20

'21

'22

'23

Graduate0 Degrees Awarded Masters

'13 '14 '15 '16 '17 '18 '19 '20 '21 '21 '23

PhD

60

Public Relations committee: Michael Beyeler (chair), William Wang, Misha Sra, Arpit Gupta, Angelina Aguilera, Michael Ray.

53

Published by the UCSB Computer Science Department.

Content and design by Michael Ray.

'24

45 30

22

15 0

2

'14

'15

'16

'17

'18

'19

'20

'21

'22

'23

Letter from the Chair Divyakant Agrawal

W

armest greetings from the tions and contributions of our highly Computer Science Departsuccessful alums prove the value of a ment to all our alums! As CS@UCSB degree well after graduwe conclude yet another academic ation. As an example of this success, year, I would like to take a moment I would like to mention that two of to reflect on the accomplishments our alums—Huican Zhu (PhD, 2000) we have made. My association with and Hongjun Zhu (PhD, 2002)—serve UCSB and the department began in as members of the UCSB Board of the late ’80s when we were still not Trustees. an independent graduate degree Given the backdrop of extreme granting department. growth in computer science, there is Fast forward to today, and our tremendous pressure on the departgraduate enrollment is the largest ment to enroll a larger number of among all academic departments on students, especially at the underthe campus. In the same vein, the graduate level. At our end, we connumber of applications for the untinue to grow our enrollment (from dergraduate computer science major 363 undergraduate students in 2014resources takes time. This is where at UCSB has increased from 2,089 in 15 to 663 in 2022-23). But this growth alumni participation becomes crucial.

2013 to 10,794 in 2023—a more For example, over the past two than quadruple increase in 10 Over the past two decades, alumni decades, alumni participation years. This rapid rise has resulted with the department has resulted participation with the department in multiple endowed chairs that in an admission rate of as low has resulted in multiple endowed as 9% compared to 30% for the have been crucial for recruiting chairs that have been crucial for campus. the best faculty especially at the A part of this growth is indeed junior level. The department is in recruiting the best faculty. driven by the demand for comneed of a similar endowment for puter science both at the underPhD fellowships to compete for graduate and graduate levels, but has to be carefully monitored and the best talent for graduate educaa significant part is also due to the measured to ensure that we do not tion. increased visibility of the department negatively impact the quality of our Finally, for our outstanding underin terms of teaching and research. pedagogy. The critical factors that graduate students, alumni engageTwo of our faculty, Chris Kruegel and contribute to the quality of teachment is crucial for Capstone projects, Tim Sherwood, were recently named ing and mentoring are the teaching summer internships, and job placeACM fellows, bringing our total num- resources: faculty members, graduments. ber of ACM fellows to seven. And our ate teaching assistants, and underIt is through this collective engagerecognition in industry and academia graduate learning assistants. Being ment that we continue to increase continues to grow as the innovaa public institution, growing these the value of your CS@UCSB degree!

CS Research Funding $15,000,000 12,000,000 9,000,000 6,000,000 3,000,000 0

2013

2014

2015

2016

2017

2018

2019

2020

2021

2022

2023 3

Agent-based Cyber Threat Intelligence and Operation

New ACTION Institute Teams Humans and AI to Fight Cyberattacks “These are some of the very best people in AI and security. They have been at the forefront of expanding the foundations of AI, machine learning, game theory, and computer security.”

Research thrusts are highly integrated and interdependent—four each in foundational AI and cybersecurity.

4

I

magine a modern city. Imagine the infrastructure that keeps it running smoothly: an aqueduct, a power plant, a smart-city system running on open-source software to monitor and distribute services. Now imagine a hostile nation-state that targets that software system, identifies vulnerabilities, gains administrative access to the main server, and shuts down the aqueduct and the power plant, causing paralysis and chaos.

The scenario is hypothetical—for now. It’s also realistic enough that the National Science Foundation recently named UCSB as the lead institution on a five-year, $20 million grant to pursue new approaches to cybersecurity.

“The basic idea of the ACTION Institute is to combine two cultures,” explains computer science professor Giovanni Vigna, the institute’s director. “One looks for new ways to do AI and the other looks to use AI in new ways to improve security.

We hope that by putting them in the same room, something amazing will result.”

The problem is complex. Countering the evolving tactics, techniques, and procedures of bad actors as they

attack massive, mind-bogglingly complicated, and rapidly changing systems is increasingly beyond the capabilities of human defenders alone. There’s just too much happening, and it’s happening too quickly.

“Solving the time-and-scale problem will require automation,” Vigna says. “And it has to be smart automation, and that means AI.” Complicating matters even further is the idea that attackers will also be using automation and AI to overcome cyberdefenses. Responding to these threats will require reasoning and acting based on small amounts of data, while adapting to untrainable and unspecified scenarios.

“In the case of security,” says UCSB CS professor Ambuj Singh, a co-PI on the project, “humans and AI may have different perspectives on the implications of actions. This is where the synthesis of humans and AI becomes useful.”

The integrated approach will use AI agents to prevent or repel attacks in time and at scale, supported by extensive domain knowledge, logic-based reasoning, and human-agent and agent-agent interactions.

According to the project proposal, over time these intelligent security agents will become increasingly robust and effective, capable of composing strategies and plans in the face of uncertainty. They will be more collaborative with each other and with humans for mutually complementary teaming, and for adapting to unfamiliar attacks.

Development work on the project is being shared among 11 academic institutions, with the goal of answering a series of research questions about the capabilities needed to build these intelligent security agents—questions covering such topics as learning and reasoning, human-agent and agent-agent interaction, and game theory.

“These are some of the very best people in AI and security,” says Vigna. “They have been at the forefront of expanding the foundations of AI, machine learning, game theory, and computer security.”

They will each be working primarily in one of eight highly integrated and interdependent research thrusts— four each in foundational AI and cybersecurity.

The ACTION Institute, in turn, is just one part of a much larger investment by the NSF to advance a cohesive national approach to AI-related opportunities and risks.

“If you use AI wrong, you can hurt entire classes of people,” explains Chris Kruegel, another ACTION Institute co-PI and UCSB professor.

“And that has occurred. So we have to be careful about what we encode in the agent’s knowledge, what we learn from the data, how we learn it, and how we align it to conform to the highest ethical values. Developing AI that is ethical and trustworthy is not an option; it’s the only thing you can do, and it’s ingrained in this community.”

“We want to have ethical AI,” adds Vigna. “We don’t want it to be making decisions that could cause harm … We want to be sure that decisions are made with a human in the loop, but in an efficient, targeted way that makes the best use of that person’s capability.”

World’s Largest Raspberry Pi Cluster Finds New Home at UCSB Computer Science What is it?

A supercomputer consisting of 1,050 Raspberry Pi 3B+. The next largest system that we know of is a cluster of 750 at Los Alamos National Lab.

Why is it?

The short answer, according to creator Chris Bensen in his series of videos detailing the project, is “because it kicks a**.” The longer answer lies in Oracle’s support for a series of outside-the-box projects designed to attract the attention of engineers and developers.

How did it get to UCSB?

Bigger on the inside: Professors Rich Wolski and Chandra Krintz with engineer Chris Bensen and his creation. to take it forward. There’s so much great technology here that they wanted to pass along for its next generation.”

How will we use it?

It started with PhD student Animesh Dangwal’s internship at Oracle Labs. “Our students are going to go crazy Professors Rich Wolski and Chanon this,” says Wolski. “We’re going dra Krintz explain that as a result to use it in a couple of ways. We of Dangwal’s outstanding work, his have large scale, high performance Oracle supervisor came to visit UCSB. computing problems that we want “In the course of discussing Anto experiment with on a low power imesh’s research, we mentioned that device. And from a high performance we do a lot with Raspberry Pis,” says computing perspective, this is a very Wolski. low power device. The other part is, They showed off their own cluster we didn’t have an instrument like this of 18 devices, and the supervisor where we can train students on the wondered if their lab would like systems end of it. This is a significant some more that Oracle wasn’t using. system.”

“I’m thinking he had maybe 20,” “It’s very different if I taught you Wolski says. “I said yeah, absolutely!” how to manage 12 Raspberry Pis,” He thought they could make room says Krintz, “than it is to do somein the shoe rack that houses their thing at this scale. This gives us the current cluster. “The next thing that ability to train students in a way that happens, here come a thousand of nobody else on the planet can.” these things.”

“You can make donuts in your kitch“This has been sitting at Oracle en,” adds Wolski, “but if you want to for two years not being used,” adds make donuts for the Super Bowl, it’s Krintz. “They really wanted someone different­—this is the Super Bowl.”

55

Neuroscience Meets AI NIH Grant Supports Development of a Smart Bionic Eye

6

Sometimes the flashes combine to form something more complex, and sometimes they don’t. It’s really a matter of understanding the neuroscience behind it—and then tapping into that as computer scientists and engineers and leading the brain into thinking it saw something.”

So the lab is focusing on the fundamentals of neuroscience, and then incorporating AI to optimize the way these devices are stimulated.

It’s a challenging project, made more so because any two people wearing the same device don’t see the same things. The Bionic Vision Lab’s team of PhD students, MS students, and undergrads is bringing together their expertise in computer science and psychological and brain sciences to explore and find solutions to these problems.

“Some of our great CS students have built deep learning models to predict—if you want to produce a certain image in the mind of the patient—how you should stimMatt Perko, UC Santa Barbara

U

CSB’s interdisciplinary Bionic Vision Lab is on a mission.

Their goal? To unravel the science behind the bionic technologies that could one day restore useful vision to people living with incurable blindness.

“One idea is to put a chip in the eye or the brain and stimulate the surviving neurons with electrical current,” explains Assistant Professor Michael Beyeler, the lab’s principal investigator. “If you do that, the other neurons can’t tell whether they were activated artificially by an electrode or naturally.”

Although the field is in its infancy, Michael Beyeler such prosthetic devices already exist.

They are known as “bionic eyes.” Beyeler’s interest is in taking them to the next level and developing smart bionic eyes.

In their current state, what the users of these devices see are basically flashes of light. “That’s what people describe,” Beyeler says. “It’s like looking at fireworks.

ulate the device. And you can personalize that strategy using feedback, like human-in-the-loop algorithms.”

For his efforts, Beyeler was recently awarded the prestigious National Institutes of Health Director’s New Innovator Award, providing a five-year, $1.5 million grant for his project “Towards a Smart Bionic Eye: AI-Powered Artificial Vision for the Treatment of Incurable Blindness.”

This is only the fifth time that this highly competitive grant has gone to UCSB. “The idea of the grant is that we can preprocess the images in a better way,” he explains. “We can use computer vision, we can use machine learning, to extract what is important and then visualize that in an intuitive way.”

As an example, he talks about looking for lost keys.

With current devices, you’re just looking through a mess of flashes of light.

“What we do is we segment the image. We would figure out, okay, here’s a table, here’s a different object on top of it, and detect in this input image where the keys are, and then highlight that for the user.”

He envisions a system that could speak with you, or you could give it audio commands. Based on your con-

“The idea of this grant is that we preprocess and segment the image. We can use computer vision, we can use machine learning, to extract what is important, and then visualize that in an intuitive way.” text or task, the demands change. If you’re in a room looking for keys, you could say “Hey bionic eye, where are my keys?” and the device would highlight them visually when you look around the room. But there’s a different set of circumstances when you’re navigating outside, for example, and you have to cross the street. In that case, the device might highlight nearby obstacles or approaching cars.

This is a big vision for a young technology, and much of the groundwork is still being laid. A number of manufacturers are working to develop devices, but none are currently close to commercialization.

“That’s one of our challenges as an academic group.

We constantly have to think on our feet, finding new partners as the industry changes.”

Some of the Bionic Vision team recently traveled to Spain to visit one such collaborator, who has a cortical device that bypasses the eye and goes directly into the visual cortex.

“We were able to test our theories and algorithms on those patients and collect data, then come home and analyze it.”

That’s just one more small but important step on a long road toward completing an ambitious vision.

New Faculty Shiyu Chang

Before joining UCSB as an assistant professor, Dr. Shiyu Chang completed both his undergraduate and PhD studies at the University of Illinois at Urbana-Champaign, followed by several years working in research labs. That includes time at the MITIBM Watson AI Lab, where he collaborated with faculty and students on advanced machine learning and natural language processing.

Dr. Chang continues this work at UCSB with research into enhancing the trustworthiness of large language models, addressing issues such as bias, uncertainty, robustness, and hallucinations. One of his highlights here so far, he says, has been teaching an undergraduate deep learning course.

“It has been incredibly rewarding to see students from various backgrounds come together, engage with the material, and bring their unique perspectives to class … Watching their growth and enthusiasm, and seeing them apply these concepts to their fields of study, has been truly rewarding.”

Santa Barbara has been a fantastic place to live and work … I feel fortunate to be part of this family.

7 7

New Faculty Dahlia Malkhi

Cutting-Edge Research in CS Supported by New Alferness-Henke Endowed Chair in Computer Science

W

Dr. Dahlia Malkhi joined the CS department at UCSB as a professor in 2024. Prior to that, her work over two decades brought scientific innovation into fruition in several leading industrial settings. Her research spans broad aspects of reliability and security of distributed systems, recently with focus on blockchains and advances in financial technology. Her work resulted in over 200 publications as well as a strong impact on computing technology.

A select sample of technologies she participated in creating includes: • Co-inventor of HotStuff, the golden standard for blockchain design, driving the Diem blockchain core engine, the Aptos blockchain core engine, and many other blockchains.

• Co-founder of VMware blockchain.

• Co-inventor of Flexible Paxos, the technology behind Log Device.

• Creator of CorfuDB, a databaseless database driving VMware’s NSX-T distributed control plane.

• Co-lead of the FairPlay project, the pioneering materialization of secure multi-party computation.

8

hen David Henke graduated from UCSB in 1978 with a degree in mathematics and highest honors, the opportunities to study computer science were limited. From the few CS courses available to him at the time, all he learned was how to program badly.

“When I took my first job after graduating from UCSB, I went to McDonnel Douglas as a mathematician-programmer,” Henke says. “It was a perfect job for me. I was doing second partial derivatives, I was doing a lot of math. We were basically taking objects and intersecting them so we could emulate and graphically represent airplanes.”

He was given nine months to program his first major project in Fortran IV. He did it in three and a half.

“I was very excited. And it all worked perfectly. My mathematician boss said ‘Beautiful work, you’ve got one error here, after that it’s perfect. Great work.’ But my programming boss said ‘This is the worst program I have ever seen. You will rewrite this program based on my standards and the standards of McDonnell Douglas or you will be fired. You have one month.’”

Adds Henke: “It was the best thing that ever happened to me—because this guy taught me how to program correctly.”

Henke, who now serves on the UCSB Board of Trustees and is an advisory board member to UCSB’s College of Engineering, would go on to have a distinguished career in computer science. After two successful startups in the 1980s, he served as director of engineering at Silicon Graphics, VP of engineering/operations at AltaVista, and senior VP of engineering/operations at both Yahoo! and LinkedIn. Computer science and computer science education have changed a lot since Henke’s own rocky start, but he remains a passionate proponent of both. At LinkedIn, which grew from 200 to 2,700 engineers during his tenure, 35 percent of those hires came straight from univer-

sities—and Henke strongly advocated for recruitment from UCSB. “I always appreciate being attached to the university,” he says.

“I give a lot of money to a lot of different things, but the number one priority is always going to be about education.”

Most recently this giving resulted in the funding of the new Alferness-Henke Endowed Chair in Computer Science, named in part for Rod Alferness, Henke’s friend and former dean of the College of Engineering.

“Rod is exactly the opposite of me,” Henke says. “He’s an introvert. He’s super smart. I’m an extrovert, not so smart. And he also would take the time to listen. That’s what I really like David Henke with his daughter on one of many visits back to UCSB. about him. He told me computer science is the place we need to invest, because it has bigger bang for the buck for not only the took him from being an engineer to being a manager. sciences, but for the whole university. Everything benefits “I didn’t want to do it. They said you really don’t have from a strong computer science department.” a choice. You either do it or you’re fired.” And in putting “The Alferness-Henke Chair in Computer Science is together a pair of teams that navigated the company the eleventh endowed chair in the department,” notes through a half-billion dollar crisis, he had a realization: “I Distinguished Professor Divyakant Agrawal, who himself learned I could get a lot more done if I had a lot of really holds the Leadership Endowed Chair good people around me … That was in Computer Science. “The availability when I transitioned from being an inof endowed chairs enables the dividual contributor to being a mandepartment to compete for the best ager. It was a hard decision. I think and the brightest talent in all areas of what you’ll find is a lot of managers computer science. We are extremely are nerds. We still like to program. grateful to David Henke for his lasting We miss it. But we’re not the center legacy to the Department of Computer of that world anymore. You have to Science and UCSB.” give that up if you want to lead other Although Henke is now retired—and people.” not for the first time—he stays inWith changes in the industry, volved in the industry as a member of though, lately he’s been considering various advisory boards and as an enanother approach. He gave his usual gaging speaker with plenty of stories leadership speech to a group of stuto tell and wisdom to share. dents recently, and afterward all they A recurring theme running through a really wanted to talk about was AI. lot of what he says is the importance of “Very rarely do I have regrets where surrounding yourself with people who I wish I could be 21 again,” he says. are smarter than you are.

“But this is one time where I think it “Big time,” he emphasizes. would be super great to be 21 and That lesson was reinforced early in to be in computer science.” his career when he started work at Silicon Graphics. Out The internet was huge, he explains. Mobile was huge. of 200 programmers at the company, he would have Big data was huge. But AI, built on top of all these things, ranked himself 197th best. is going to be even bigger. It’s going to disrupt the way “I’ll tell you, that was a hard pill to swallow,” he says. things are done. It’s going to require retraining for a lot “But it was a good one. I figured either I’m going to feel of currently comfortable people. But it’s coming. bad about this or I’m going to feel good. And the good “If you’re not there doing it, you’re a dinosaur,” he part is I can learn. I can learn from these others.” warns. But he sees it as an exciting time. “If I could It was that attitude combined with drive and dedicabe there as a young engineer training these systems, I tion—and a bit of arm-twisting from the company—that would love it.”

Computer science is the place we need to invest. It has a bigger bang for the buck not only for the sciences, but for the whole university.

9

Capstone 2024 The Capstone course sequence presents an opportunity to develop innovative solutions to real industry problems.

Y

ou won’t find answers to these problems in textbooks, on message boards, or in off-the-shelf software packages —because students in this class are among the first ever to solve them. Working together with industry leaders, student teams take on the most challenging problems of the day with technological innovation, creativity, and boatloads of hard work.

Led by professors Tevfik Bultan and Giovanni Vigna, nine teams of students this year united with industry leaders from Artera, Appfolio, Terawe, Forta, Aziksa, and Veridise. The challenges they addressed ranged from making it easier for non-native English speakers to communicate in the workplace to using AI for early detection of melanoma.

10

Capstone courses are offered during fall and winter quarters at UCSB, culminating in project presentations before industry judges in mid-March at Henley Hall.

Capstone comprises two courses: CS189A and CS189B. In the first course, students choose a project and develop it with milestones such as vision statements, requirements, and design goals. The course culminates with inclass presentations and demos. In the subsequent course, students develop real systems for their projects, test them in front of real users, adjust their designs based on feedback received, and finally show off their projects publicly in a day of poster presentations, lunch, one-onone talks with industry judges, and final presentations ­—with cash prizes going to the top three teams!

The Capstone program is a fun and exciting opportunity to gain practical experience and connections, making it a highlight of the CS undergrad experience at UCSB!

Winning Teams First Place: Team Realty Reel. A platform for creating professional videos of real estate properties in minutes using AI and machine learning.

Second Place: Team Aware. An Apple Watch alcohol risk estimator to determine levels of intoxication by measuring walking steadiness and heart rate.

Third Place: Team Logos. A universal platform for developing, compiling, and auditing ZK circuits.

New Faculty Wenbo Guo

Dr. Wenbo Guo joined UCSB as an assistant professor after completing his postdoc at UC Berkeley, his PhD at Penn State, and his master’s at Shanghai Jiao Tong University.

His interests center around the intersection of computer security and machine learning.

Recent research has focused on designing foundation models and building Reinforcement Learning-driven planning systems for software and network security problems, and improving the explainability and robustness of large models and RL. He was drawn to UCSB for its strength in security and ML, specifically noting the ACTION Institute and the Center for Responsible Machine Learning.

“Our faculty are collegial, talented, and nice individuals,” he says. “I enjoy collaborating with them.”

In his free time, Wenbo enjoys playing and watching sports, and he finds that Santa Barbara is an ideal place for pursuing that passion—not to mention being an ideal place for CS research!

Students here are of high quality and working with them is enjoyable — and of course the campus is gorgeous.

11 11

New CS Faculty Welcome Party Divy and Shubra Agrawal opened their home this spring to faculty and friends for an event welcoming the four newest members of the Computer Science Department faculty family.

Murphy Niu, Dahlia Malkhi, Wenbo Guo, and Maryam Majedi (pictured above) braved a windy Santa Barbara sunset to join with seasoned faculty, staff, partners, and CS supporters for an evening of food, music, and fun.

Professor Tevfik Bultan entertained by bringing (some of) his band Previously Committed, including Sumita Pennathur (professor of mechanical engineering) on alto sax and Aashish Mehta (professor of global studies) on keyboards and flute. Bultan’s son Aydin joined on drums.

Delicious Indian food from Masala Spice kept everyone warm as the sun set, and then the ice cream got them all dancing—led by the smooth moves of Professor Giovanni Vigna and Professor Agrawal himself!

Viewers of Instagram Reels were clearly envious of all the fun when they wrote “Nothing like🔥 Bollywood hits to get a CS party going!” and “More turnt than IV on a Friday night! ”

12 12

13

New Faculty Maryam Majedi

Broadening Participation in Computing with Associate Teaching Professor Richert Wang The National Science Foundation has awarded a $300,000 grant for a demonstration project led by UCSB’s Richert Wang and CSU Chico’s Kevin Buffardi.

The project focuses on improving representation, relevance, and equity in computer science by framing high school instructional practices from the perspectives of women and Hispanic students.

I Dr. Maryam Majedi joined UCSB in July 2023 as an assistant teaching professor. She holds a PhD in data privacy from the University of Calgary and an MSc in high-performance scientific computing from the University of New Brunswick, as well as a fellowship in medical innovation from Western University.

Her research falls into the intertwined areas of data privacy and embedded ethics. She looks forward to teaching the next generation of scientists, educators, and technology developers how to incorporate ethical considerations into computer science technologies throughout their careers.

“My first impressions of UCSB have been very positive,” she says. “The department feels welcoming and dynamic, with collaboration between students and professors. The passion for learning and research is evident, and it is clear that the department values innovation and wants everyone to feel involved.”

UCSB stands out as an exceptional institution. I treasure every moment spent here.

14

n CS literature, there’s a lot of evidence showing that retention rates are directly correlated to a lack of identity. When students are in high school and they don’t see people like themselves in academic or professional positions, for example as software engineers, it sends an early message that this is something unattainable for them.

“The field of computer science does struggle with representation,” explains UCSB Associate Teaching Professor Richert Wang. And that’s a problem, because industry is eager to hire people from diverse backgrounds with diverse perspectives. “It’s all about inclusive design.

We need people to think differently in order to build the best products possible.” He gives examples of crash test dummies modeled after male engineers, resulting in higher casualty rates for female passengers, or digital cameras that turned out to have difficulty focusing on darker skin. “Just having that representation, that different point of view, really helps improve products in the industry.”

But bringing these diverse voices into computer science presents plenty of challenges. Lots of students, especially from lower income school districts, have never had the opportunity to study CS in high school. And once they get to college, CS courses can be very difficult to get into. And if they manage to get that far?

“Imagine somebody coming in, facing impostor syndrome, seeing a bunch of people that don’t look like them already seemingly knowing what to do—that is really discouraging. Having role models is essential for keeping students in the classroom, keeping students interested, and having students believe that they can actually do computer science.”

With their Broadening Participation in Computing project, Wang and Buffardi are attempting to broaden the pipeline into CS by opening doors at the high school level. The pilot project is a platform of instructional material with auto-graded exercises. It will be tested next year,

“It’s all about relatability.

It’s about mentoring student peers to introduce concepts, serve as role models, and Office hours with Richert Wang, recipient of UCSB’s 2023-24 Distinguished talk about Teaching Award. computer tive of computer science than what including locally at Santa Barbara’s you typically find in a textbook.”

Dos Pueblos High School, and evenscience in their That means things like teachtually it will be available for anyone ing object-oriented programming to access and use for free online. own words.” concepts in the context of In-n-Out’s Wang is currently mentoring three CS undergrads in the creation of instructional videos covering various curriculum topics. “We want these students to not only teach computer science, but have the audience and prospective students get to know them as people and role models.

So the examples they’re coming up with are relatable to their interests, relatable to their life, and kind of provide their own unique perspec-

secret menu, or learning about multidimensional arrays as a canvas for pixel art.

“Even if a student is not interested in computer science, they can use this as a learning tool to get some sort of informal instruction, to see some relatability of peer mentors.

The hopes are this will open some doors that weren’t open for them before.”

15

Human-AI Integration Lab Human Augmentation with Interactive AI + XR

“T

his is a busy, busy place” says Assistant Professor Misha Sra, director of UCSB’s Human-AI Integration Lab, also known as HAL. “It’s not just PhD students. I have a whole lot of undergrad students, I have masters students, I have visiting researchers like Kojiro who’s here from Toyota Research. I also have faculty spending their sabbatical year here from Japan. It’s a lively place.”

And working diligently in all that buzz of activity for the last few years has been PhD student Atieh Taheri, whose lifetime of experience with spinal muscular atrophy has broadened the perspective of a lab that’s focused on human-computer interaction, extended reality, and AI. “My field of research is accessibility for people with motor impairments, motor disability,” Taheri says. “People kind of similar with the situation I have.”

As an example she talks about current virtual reality implementations, and how they mostly consider the

general population while overlooking people in different situations or medical conditions.

“Whatever experience you’re trying to simulate for people,” she says, “you also need to consider this underrepresented group.”

“This has been the challenge in Atieh’s prior work,” Sra explains. “When you think about gaming systems and input devices we have for computers, everything assumes that you can use your hands for the most part. If you don’t have control over your hands, accessibility features are not going to help you.”

With that in mind, Taheri’s first project in the lab was a gaming input system based on facial expression recognition. She built it during the Covid shutdown, conducting a remote user study with 12 other people who have muscular dystrophy or spinal muscular atrophy.

“It was one of the hardest studies I have ever seen done,” Sra says. “Atieh did it, building a system that was

“Well I mean she’s an engineer, right? And that’s what engineers do. They solve hard problems.”

PhD candidate Atieh Taheri with Assistant Professor Misha Sra, director of the Human-AI Integration Lab.

16

usable for folks who can’t necessarily rely on their hands to set up and play.”

The system, which relies on AI and a web camera with speech input, is customizable by letting users with different degrees of facial and muscle control map different

• ModBand ’23 • Sonic Storyteller ’23 • ARFit ’23 • Appearance Editing ’23 • FaraPy ’21

New Faculty Ziad Matni

• EntangleVR ’23 • ARLang ’23 • Spatial Music ’23 • Emotion Awareness ’23 • Spatial Memories ’23 • ARfy ’23 • Experiential Artifacts ’22 • CardsVR ’22 • Virtual Steps ’24

• TiNO-Edit ’24 • DanceGen ’24 • GITM ’23 • Living Memories ’23 • Virtual Instructors ’23 • EChat ’23 • Recoloring ’23 • Mix3D ’22 • Txt2Vid ’22 • GenAI ’22

• Mouse Clicker ’24 • Hands-free Input ’21

• Virtual Buddy ’23 • Facial Expressions ’21

Visual summary of recent work in the Human-AI Integration Lab (HAL) facial expressions to game character motions or actions.

Games downloaded from Steam that allow keyboard remapping worked well right out of the box with the system.

“We were going for Atieh to play a first-person shooter.

And she did! Car racing, too. Hard games!”

Sounds difficult?

“Well I mean she’s an engineer, right?” says Sra. “And that’s what engineers do. They solve hard problems.”

During her time in the lab, Taheri has also worked on a computer mouse modification project that provides tactile feedback for people with hand motor impairments, a virtual buddy for redefining conversational AI interactions, and Virtual Steps, a VR experience that simulates walking for lifelong wheelchair users.

“I don’t know if my walking in VR is the same as real walking in physical reality, I don’t know that,” Taheri says.

“But it’s comfortable for me and it’s consistent with the model I have in my mind from watching others. It’s consistent with my mental model of walking.”

She’s quick to point out that not everyone will want to experience walking—and that’s fine. Her projects are more about bringing focus to those with disabilities, about bringing awareness to designers who don’t necessarily have the perspective of disability. Sra adds that disability is not necessarily a siloed condition. It’s more of a continuum, with people sliding in and out of degrees of disability throughout their lives with illness, injury, and age—making these kinds of projects relevant to all of us.

As a Presidential Postdoc Fellow, Taheri will continue to expand her knowledge and expertise in accessibility research while contributing to the vibrant research community at Carnegie Mellon University.

Meanwhile, the lively work continues in the Human-AI Integration Lab as new waves of researchers innovate and expand the possibilities of human augmentation with AI and XR.

After earning his MS in EE/CE from USC, Dr. Ziad Matni took a short 13-year break in the tech industry before returning to academia for a PhD from Rutgers. He joined UCSB as an assistant teaching professor and now serves as director of the CS department’s Early Research Scholars Program.

His own research focuses on the intersection of computer science and human behavior around information seeking and sensemaking—including the use of AI and Large Language Models in education. He loves teaching undergrads but—after recently getting nothing but blank stares from his very bright students upon mentioning the names Jobs and Wozniak—he now fantasizes about adding a history of CS course!

“UCSB is the best of many worlds,” he says. “It’s a world class research leader in a gorgeous location. I’ve found the faculty to be intellectually stimulating and very approachable.

The staff consistently demonstrates exceptional generosity with their time and expertise. And our students are sharp and eager to learn.”

Our students are sharp and eager to learn.

They really are a pleasure to teach and mentor.

17 17

New Faculty Murphy Niu

CS Department Debuts New Initiative to Aid Underrepresented PhD Applicants

T Dr. Murphy Niu describes herself as a physicist by training and a computer scientist by choice. As an assistant professor at UCSB, her research deals with harnessing the imperfections and the unique advantages of different quantum hardware through machine learning and algorithm designs.

She looks forward to introducing a new undergraduate course that will teach paradigms in quantum mechanics and quantum computing to CS students with no background in quantum physics.

Dr. Niu is excited by the thriving quantum research community at UCSB but also by the campus culture and the beautiful natural environment.

“I’ve visited Santa Barbara often,” she says, “but nothing compares to actually living here.” It’s a perfect place for some of her favorite activities old and new, including bird watching, wildlife advocacy, hiking, scuba diving—and anything else that will keep the learning curve high.

There’s so much to learn, observe, listen to, and enjoy here in Santa Barbara. I’m just getting started!

18

he Computer Science Department’s committee on Diversity, Equity, and Inclusion has announced the launch of the PhD Application Support Program. Spearheaded by Professors Michael Beyeler and Frederic Gibou, this initiative supports aspiring CS scholars from under-served backgrounds through a preapplication review service.

The PhD ASP provides mentorship and feedback on application materials, including resumes, CVs, and statements of purpose. Current PhD students or postdocs from the CS department serve as reviewers, undergoing specialized training to offer insightful advice.

In its first year, the program supported 53 students, mostly first-generation college students, from 43 institutions in 11 countries. An anonymous survey revealed participants found the program invaluable, praising the feedback for significantly improving their application materials. Each application receives two reviews for diverse perspectives. One participant noted, “My materials were read carefully, and very useful suggestions were provided. I am extremely grateful for this support program.” The program aims to expand in the coming years to benefit more students.

Prospective applicants from underrepresented or marginalized backgrounds in CS are encouraged to submit their application documents for review. The program guarantees confidentiality, ensuring that participation details are not shared with the faculty committee reviewing applications.

Recognizing the systemic challenges faced by many applicants from marginalized communities, the program emphasizes guidance and improvement. “It comes with no guarantee of admission, of course, but is meant to strengthen the written materials of those who might ‘not look so good on paper’,” noted Professor Beyeler.

We hope to start the next cycle of the PhD ASP in October ’24. At that time, interested prospective students will be able to apply at https://cs.ucsb.edu/ education/graduate/phd-application-support-program.

By introducing the PhD Application Support Program, CS@UCSB reaffirms its commitment to fostering diversity and inclusivity, providing valuable support to prospective scholars.

Big thanks to our outstanding CS staff for all they do to keep the department running smoothly every day!

19

Awards & Honors Congratulations to our faculty, undergrads, grad students, and alumni who continue to impress with notable successes on and off campus. Your hard work and dedication elevates the profile of UCSB Computer Science and collectively increases the value of a CS@UCSB degree!

Divyakant Agrawal

Prabhanjan Ananth

IEEE MDM 2024 Test of Time Award NSDI 24 Outstanding Paper Award

NSF Early CAREER Award

Jonathan Balkind

Elizabeth Belding

Michael Beyeler

Amr El Abbadi

Yu Feng

NSF Early CAREER Award

Best (Long) Paper Award, ACM Internet Measurement Conference

NIH New Innovator Award

IEEE MDM 2024 Test of Time Award NSDI 24 Outstanding Paper Award

Distinguished Paper Award, ACM SIGPLAN Conference on Programming Language Design and Implementation

Arpit Gupta

Tobias Höllerer

Chris Kruegel

Diba Mirza

Timothy Sherwood

Best (Long) Paper Award, ACM Internet Measurement Conference Applied Networking Research Prize, IRTF Best Paper Honorable Mention Award, ACM CCS

Induction to the IEEE VGTC VR Academy

ACM Fellow

2022-23 CS Department Outstanding Faculty of the Year Award

ACM Fellow

Misha Sra

Richert Wang

William Wang

Lingqi Yan

Xifeng Yan

2023-24 Hellman Fellowship NSF Early CAREER Award

NSF grant for Broadening Participation in Computing UCSB Academic Senate Distinguished Teaching Award

Pierre-Simon Laplace Early Career Technical Achievement Award CRA-E Undergraduate Research Faculty Mentoring Award Karen Spärck Jones Award

Best Paper Honorable Mention, ACM SIGGRAPH

VLDB Test of Time Award

20

2024 Distinguished Alumni

John Yoder, undergrad Maureen Heymans, MS Daniel Nurmi, PhD

College of Engineering Outstanding Teaching Assistant

Junrui Liu

Computer Science Outstanding TA Awards

Vinayak Gajjewar, Trysten Hess, Galen Pogoncheff Computer Science Outstanding Publication Award

Xinyi Wang

Computer Science Outstanding Mentoring Awards

GauchoChat Team: Rajan Saini, Weizhi Wang, Marina Zhukova, Hong Wang, and advisor Xifeng Yan

Arthur Caetano, Radha Kumaran Computer Science Outstanding Dissertation Awards

Ishtiyaque Ahmad, Tsu-Jui Fu

Caitlin Scarberry Memorial Award: Computer Science BS/MS Students of the Year

Erwan Fraisse, Aditya Sharma

Computer Science MS Student of the Year

Ryan Keeney

Computer Science PhD Student of the Year

Zach Sisco

Alexa Prize SocialBot Grand Challenge 5

GauchoChat Team, First Place Alexa Prize SimBot Challenge

GauchoAI Team, Second Place NSF Graduate Research Fellows

Haarika Manda, Joyce Passananti CRA Outstanding Undergraduate Researcher

Isaac Hair, Runner Up Matthew Ho, Honorable Mention Shivani Sista, Honorable Mention

Goldwater Scholarship

Rising Stars in Machine Learning Awarded by the University of Maryland

NSF Graduate Research Fellowship Program Honorable Mention

Wanrong Zhu

Isaac Hair

Guy Wilks

UCSB Chancellor’s Award for Excellence in Undergraduate Research

Guy Wilks

College of Engineering Outstanding Senior Awards

Jonathan Cheng, Arik Cheslog, Andy Ouyang, Ivy Wang Computer Science Outstanding Undergraduate Mentor Award

Tianle Yu

Computer Science Outstanding Undergraduate Research Award

Isaac Hair

Computer Science Students of the Year

Vaishnavi Himakunthala, Zackary Glazewski

IJCNLP-AACL Area Chair’s Award

Liangming Pan

WG 2024 50th Intl Workshop on Graph Theoretical Concepts in Computer Science

Ryan Keeney, Best Student Paper Award

UCSB 25th Annual New Venture Competition

Bowen Zhang (EyeClimate Team), First place CSAW Logic Locking Competition 2023 at NYU

Yeganeh Aghamohammadibonab & Jaber Daneshamooz, Second Place

UCSB AI Community of Practice Spring Symposium 2024, AI Digital Presentation Competition

Tianle Yu, First Place

Neal Fenzi Resonant Founder Fellowshp

Michael Saxon, Vaishali Surianarayanan, Zach Sisco 21

Get Involved

T

he Department of Computer Science at UC Santa Barbara is a world leader in computing research and education.

Our highly ranked graduate and undergraduate programs are in high demand and attract ambitious and talented students. Our faculty and students produce some of the finest research in the field, and our alumni in academia and industry create solutions for the world’s most challenging problems.

The continuing support of our alumni and friends is critical to our ability to grow and build on our success. The department’s priority funding areas provide a number of ways to impact our future.

Student Support Through the Student Support Fund you will enhance the overall student experience and ensure that UCSB Computer Science continues to attract and educate the students who will become tomorrow’s leaders.

• Named Scholarships & Fellowships • Named Lecture Series

• Undergraduate Lecture Series • Distinguished Lecture Series • Unrestricted Support

Research Support Research-directed gifts help us recruit and retain the very best faculty and graduate students from around the world who will focus their talents and energies on creating technological solutions to increasingly complex global problems. These gifts help us remain competitive with our peers and continue to elevate our national and international stature.

• Named Endowed Chairs

• Named Visiting Professorships

Capital Projects UCSB Computer Science students and faculty need facilities and spaces that support collaboration, inspiration, and discovery. The Department seeks support of a generous donor to establish a new state-of-the-art building.

Smaller scale gifts are sought to help upgrade and maintain facilities and to support priority capital projects, including new expanded instructional computer labs for our undergraduate students.

Corporate Partnerships tinyurl.com/muvrhajw 22

Today’s UCSB Computer Science students are the future software engineers, researchers, and architects who will drive your organization into the next generation of technology and beyond. Consider taking the first step to invest in your future. Check out the UCSB Corporate Relations page for more information: http://tinyurl.com/35nw2kb8

Special Thanks Ryan M. Allen ’18,’19 Dr. Amr El Abbadi and Janet Head Nayan Ashar Richard L. Basanese ’86 Gang Bao Julie Bechtloff ’95 and George Weising Bright Funds Omar K. Buzdar ’23 Wai Chan and Justin Liu Dr. Douglas Chang ’98 Eunice Cheung Jackie and Raymond Choi Dr. Hae Don Chon ’02 Sharon Clay Carolyn A. Cox ’88 Jennifer Rose Cryan ’15,’17 Gregory E. Davis ’98 Mayumi ’83 and Dr. Gregory DeHaan ’83,’85 Lei Dong and Changlin Ma Dr. Zoran Dimitrijevic ’04 Suvanan Dutta Daniel ’02 and Rebecca Elder ’02 Ethereum Foundation Lopker Family Foundation Floyd Goldstein Dr. Nevena Golubovic ’19 Paul D. Goodrich ’95 Mayumi and Roy Gottlieb Larry D. Hale and Yanmin Tao Chris and Jean Hamabe Kelly Harrel Karin E. Havemose ’88,’91 Yi He and Rong Feng Hannah Hultgren David Henke ’78 and Elana Tabachnikoff Deanna M. Hartsook ’16 Suzan D. Jackson ’86 Xiangdong Jin and Aihua Fu Sara D. Kalin ’97 Saagar S. Kausheeka ’18 Vach and Jacintha Kompella Forrest P. Laureano ’67 Brian H. Lee ’12

Dr. Kuo-Chin Lien ’16 The Susan & Bruce Worster Google, Inc. Dr. Rone Kwei Lim ’12 Foundation IBM Corporation Dr. Honglei Liu ’17 Chen Wu and Yanmu Huang Intel Corporation Matthew M. Misiurak ’06 Daniel Wu ’04 Invoca, Inc. Fereshteh ’99 and Neil Morgan Dr. Ping Wu ’07, ’08 and Juan JPMorgan Chase Bank, ’99 Wang ’08 N.A. John and Janis Mori Xiujun Wu Linctex Digital Technology Jana and Jonathan Multhaup Jenny Xia and Dennis Fang LogMeIn, Inc. Dr. Priya A. Nagpurkar ’07 Ping Yang MAZ Group NCWIT Xiaodong Yang and Yu Fang McNally Media Rong Nie and Qing Zhou George Yuan Megagon Labs Open Source Hardware Chunlian Zhang and Gang Lu Microsoft Association Zhongdang Zhan and Rong MystenLabs Ramesh Pabbichetty and Vakula Zhang Nanjing Weiwu Internet Vura Daping Zhang and Zonghuan Technology Co.,Ltd Lucien and Jessica Patenaude Wu NEC Laboratories America, Kimberly and Carl Pawley Yihong Zhang and Benquan Yu Inc. Conor M. Pawley ’11 Guoxing Zhao and Lusheng PayJunction Patricia and Jim Pflaging Chen PicsArt, Inc. Linda Rice-Zamora Jingyi Zhong and Yifeng Yang Pivotal Ventures Michael J. Richmond ’05 Qiyu Zhong, Class of 2024 PowWow Energy Pedro Rivas Hongjun Zhu ’98 and PricewaterhouseCoopers, The Rockefeller Foundation Jing Wang ’98 (Ann Jie LLP Dr. Onur Sakarya ’08 Foundation) QAD, Inc. Suzanne and Mark Scarberry Jun Xiu Zhu and Jian Wen Zeng Qualcomm, Inc. Bill and Christina Sequeira Zhu Family Foundation, Inc. Samsung Semiconductor, Amy Shen and Robert Jiang AbbVie Inc. Xiaodong Shen ’99 and Jane Xu Activision Publishing, Inc. Schwab Fund for Charitable Mel Sibony ’79 Adobe Systems, Inc. Giving Renee Slattery ’87 Alcon ScOp Ventures, LLC Joel Silverman Algorand Semiotic AI Soobin and Christian Son Allthenticate Shellphish Support Liling Song ’02 Amateur Radio Digital Syndicate Tayloe and Cindy Stansbury Communications SmartContract, Inc. Sanjay and Ashima Suri Amazon Web Services, SmartRG Inc. Jufang Tan Inc. Snap Inc. Dr. Hong Tang ’04 and Man Luo Amazon.com, Inc. Snowflake, Inc. Qun Tang Ampere Computing LLC SyncReality Jerry and Wendy Tsai AppFolio, Inc. Teladoc Health Tentcho Totev Artera Telekom Colleen Vandevoorde Beegol TrueVision Systems, Inc. Matthew ’06 and Sara ’07 Beijing ByteDance Technology Viasat, Inc. Vlasach Co., Ltd. Visa, Inc. Chengbo and Xiaohua Wang Benevity Lam Research Watermark Insights, LLC Liaoteng Wang and Yulin Zhang Benevity Pfizer www.Bill.com Randy Wang ’13 Cisco Systems, Inc. Xenon Partners Zhiqiang Wang Digilent XVerse Technology Waterclowns Evidation Health, Inc. Yardi Systems, Inc.

Retirement Update: Ömer Eğecioğlu Computer Science Professor Emeritus Ömer Eğecioğlu published a new book in 2023 titled Fibonacci Cubes with Applications and Variations with his coauthors Sandi Klavžar (Ljubliana) and Michel Mollard (Grenoble). This compendium features the state of research into Fibonacci Cubes, which have been a popular area of inquiry since the 1990s. The book serves as a fundamental source for further research, benefiting advanced students in CS and mathematics.

23

College of Engineering Santa Barbara, CA 93106-5130

UCSB Computer Science thrives with the active, enthusiastic support of our many alumni and friends.

Your contribution makes a difference to students and researchers as they create solutions to the world’s most challenging problems.

tinyurl.com/muvrhajw 24