A group photo of USC SAIL lab’s director, alumni, current students, and guests at its 25th anniversary celebration. (Credit: David Chou)
This winter, a laboratory reunion held at USC’s Ming Hsieh Institute (MHI) lounge transformed into a full-scale international engineering conference, featuring technical panels, talks, and poster sessions highlighting ongoing research projects.
Hundreds of SAIL alumni returned, and research collaborators, industry leaders and students from around the world gathered on Jan. 16-17 to celebrate the 25th anniversary of the Signal Analysis and Interpretation Laboratory (SAIL). Housed under USC Viterbi School of Engineering’s Ming Hsieh Department of Electrical and Computer Engineering and School of Advanced Computing, the lab specializes in signal processing and speech emotion recognition. Founded by USC University Professor Shrikanth “Shri” Narayanan in 2000, the SAIL lab is dedicated to addressing key social issues and developing human-centered technologies that help the world understand human conditions, and ultimately better the human experience.
Today, SAIL has evolved into a global leader in speech analysis and building mental health technology, with expanded interdisciplinary work in areas like human health and machine intelligence, as well as a growing emphasis on artificial intelligence (AI) as the lab continues to evaluate its priorities to meet changing demands around the world and shifting industry landscapes.
A Trailblazer in Human-centered Innovations
“SAIL is all about creating technology that puts ‘human’ at the forefront in every part of the work we do,” says Narayanan’s PhD student Kleanthis Avramidis as he explains the lab’s mission.
The lab’s success goes beyond the number of papers published, patented technologies and award-winning projects.
From being a pioneer in the speech recognition field to contributing key data that laid the groundwork for research, and developing countless tools that solve pressing real-world problems, the scale of the lab’s innovations spans across academia and the research field but also our everyday lives.
Described as “ahead of its time” by Albeer Alwan, professor at UCLA, Narayanan’s mentor and PhD advisor, SAIL is a trailblazer in many research areas, especially speech emotion recognition and real-time magnetic resonance imaging (MRI) for speech production.
SAIL was one of the firsts to observe how speech is produced as Narayanan pioneered the use of real-time MRI imaging to visualize the dynamic movements of the human vocal tract during speech, allowing researchers to see exactly how humans create sound, leading to critical breakthroughs in understanding language acquisition, beatboxing and clinical vocal tract disorders.
Bridging the gap between engineering and education, Narayanan also spearheaded a project with Alwan and other researchers from UC Berkeley using speech recognition to automate literacy assessments for children. These tools were designed to facilitate personalized education, particularly for younger children, and help teachers provide personalized instruction, as well as track student progress more effectively. More than 15 years ago, the findings from this project was among the firsts to tackle challenges in acoustic variability with speech recognition and inclusive technologies.
Datasets that Powers the Future of Research
Today, SAIL’s papers have been cited more than 5,000 times, with the widely used IEMOCAP database remains a highly influential resource.
“The data we collected over the decades is SAIL’s defining contribution to the research world,” said Kleanthis. Antonio Ortega, the lab’s close collaborator and professor at USC Viterbi School of Engineering’s Ming Hsieh Department of Electrical and Computer Engineering and School of Advanced Computing, also echoed that sentiment in an interview, emphasizing the lasting impact of the lab’s datasets on the broader research community.
With these massive, intricate datasets, SAIL plays a key role in laying the foundation for researchers across linguistics, clinical psychology, medicine, children’s education and media studies.
Many scientists use these data to further their research and power both academic and industrial innovations. A prime example is the data from the Tracking Individual Performance with Sensors study (TILES), where the lab collected physiological signals from nurses at Keck Hospital of USC to identify markers of workplace stress. The data is later being applied to create wearable technology and AI-driven models to help office workers with demanding jobs manage stress.
“A lot of this data is not easy to collect,” Dani Byrd, a close collaborator of SAIL lab and USC Dornsife linguistics professor, further pointed out.
SAIL didn’t just collect data– the lab did the heavy lifting to process difficult datasets SAIL alumnus Kartik Audhkhasi, and current research scientist at Meta’s Superintelligence Labs, also added that it was a particularly challenging task, as “much of this data is very noisy and hard to clean up since they are collected from real world scenarios, and not controlled environments with less variability.”
Currently building robust speech recognition systems that aims to accurately interpret diverse human accents and dialects at Meta, Audhkhasi believes this kind of work is hard but necessary, “as real-world research problems that we try to solve often do not involve clean data.”
Technologies That Change Lives
Beyond datasets, SAIL has continuously offered solutions to a variety of social causes that have directly changed lives over the years, from addressing mental health challenges to advancing more inclusive media.
In the area of mental health, SAIL developed Behavioral Signal Processing framework to address social causes such as autism, addiction, depression, suicidality and couples therapy. The lab develops computational methods to “illuminate behavioral traits and mental states,” using AI to analyze human behavior from signals such as speech nuances, heart rate and even skin conductance to support clinical research. The BSP team investigates multimodal behavioral signals to build computational models of human behavior and create informatics-driven scientific discovery and decision-making.
Using these signals, the SAIL team designed tools and technologies that automate the evaluation of addiction counseling, track emotions from speech and body language, detect depression and suicidal ideation to support clinical screening, and classify and predict outcomes in marriage counseling based on speech and interaction patterns, including audio recordings from couples in therapy.
SAIL has also addressed media bias as a social cause, particularly gender bias in media language with its Computational Media Intelligence (CMI) work. A primary goal for the lab is building technologies for inclusion. These CMI tools use AI analysis to examine 2.7 million advertisements to quantify gender representation and identify biases in film and television, analyzing video, audio and text, and focusing on metrics such as on-screen time, speaking time and character agency. The project was conducted in collaboration with the Geena Davis Institute on Gender in Media.
A Multigenerational Family Tree
With human-centric research as the lab’s core principle, people are at the heart of SAIL—from the founder himself to current students and the lab’s alumni.
Many alumni and students described the people of the lab as what makes interdisciplinary research, a global web of professional networks and a generational research family possible, citing the tight connections across the growing generations of SAILers.
Narayanan values mentorship and fosters a supportive lab culture, empowering students to lead projects, take on challenges and work in a collaborative safe space to generate ideas as a team.
“A very precious, unique side of the lab is the interdisciplinary aspect of SAIL,” noted Kartik, with many alumni and research collaborators also describing that interdisciplinary is what sets SAIL apart.
This lab culture sparks discussions of cross-area collaborations within the lab and promotes partnerships across different research disciplines, which ultimately contributes to SAIL’s success in creating a breadth of research outcomes and generating joint projects across disciplines within USC campus— spanning linguistics, Keck Medicine of USC and psychology, and beyond.
This people-first culture has also inspired new labs and talent around the world, as alumni establish research groups that replicate SAIL’s values and environment. SAIL alumnus Carlos Busso, for instance, is now a faculty member at Carnegie Mellon University and founder of the Multimodal Speech Processing Laboratory. Deeply inspired by his personal experience at SAIL, he built a lab focusing on collaboration and mentorship.
“The best part about SAIL is the family,” explained Huang-Cheng (David) Chou, Narayanan’s current postdoc student and also Busso’s former visiting PhD student at his lab. “Shri is technically my academic dad but also my academic grandpa, isn’t that cool.”
SAIL is also a home away from home that many return to. The group builds a vibrant social life and promotes human connection among lab mates through social events and lab traditions, encouraging SAILers to “have fun,” become well-rounded researchers and grow beyond the confines of the laboratory.
Events such as conference meetup traditions, annual social gatherings, and summer music parties, are what have strengthened connections, even long after students graduate.
Over the past 25 years, the lab has experienced exponential growth in size, with hundreds of alumni now working in leadership positions at top technology companies and research institutions such as Meta, Apple and Georgia Institute of Technology, among others.
Today, SAIL is known for its wide global professional network across industry and academia, as generations of SAILers have formed a powerful and expansive web of connections. This growth continues to attract talent and students. Many alumni note that they have benefited from the “SAIL network” throughout their careers, and current students describe this network and professional development opportunities as top reasons for joining the lab.
The Legend Behind the SAIL Legacy
While the research and community reflect the lab’s achievements, SAIL’s founder and director is the internal compass driving the growth.
Narayanan has shaped SAIL into an idea hub where students lead their own projects and pursue their own creative initiatives. This leadership style has turned the lab into a professional engine that transcends a typical advisor-student relationship, creating a lifelong “family” bond.
A group photo of Narayanan (top left), his PhD Advisor Abeer Alwan, (center), and lab mates.
Alwan explained that SAIL’s global success today is rooted in Narayanan’s perseverance and resilience that started when Narayanan was still a student. As her first PhD student, she recalls that he would race to Cedars-Sinai at any time of the day just to catch cancelled MRI slots to observe human vocal tract since the lab had limited resources as a new faculty. “It was this type of dedication and ambition spirit in doing research that led to all the breakthroughs”, Alwan added.
As Narayanan mentor, Alwan describes Narayanan as “Creative, multidisciplinary, and visionary”. She notes that he has always chosen projects with “great social impact,” another defining trait of SAIL’s core mission.
As Narayanan’s long-time collaborator at USC Dornsife’s linguistics department and former classmate at UCLA, Byrd further highlights his success as a scientist is rooted in his “unique take on science,” incredibly optimistic character and “very open-minded” attitude approaching technical challenges.
To Alwan, seeing generations of students and meeting her “academic grandchildren” gathered to celebrate the lab at this year’s anniversary event is a testament to the SAIL’s human-centered values that Narayanan has cultivated since day one.
This year, Narayanan was inducted into the National Academy of Engineering, marking a significant milestone and achievement for both himself and SAIL, coinciding with the lab’s quarter-century anniversary. Narayanan, the Niki and Max Nikias Chair in Engineering, also holds faculty appointments in USC’s Dana and David Dornsife College of Letters, Arts and Sciences and the Keck School of Medicine, and continues to serve as USC’s inaugural Vice President for Presidential Initiatives. Holding multiple leadership positions, he continues to serve as USC’s inaugural Vice President for Presidential Initiatives and the MHI Director.
Learn more about USC SAIL’s current and past research projects here.
Published on March 25th, 2026
Last updated on March 25th, 2026