Joseph P. Campbell
Dr. Joseph P. Campbell joined MIT Lincoln Laboratory (MIT LL) in 2001 as senior staff in the Information Systems Technology Group. He became assistant leader, associate leader, and leader of this group, renamed Human Language Technology (HLT), in 2010, 2011, and 2016, respectively. In 2019, the group became Artificial Intelligence Technology and Systems, and Campbell began rearchitecting it to continue creating world-class HLT and advance deep learning methods for cyber analytics and counterinfluence operations. He specializes in developing and transferring AI technologies for government applications and operationally relevant evaluation to achieve tremendous mission impact.
Early on, Campbell led the development of forensic speaker-comparison technology for the U.S. Secret Service, MIT LL's first U.S. Department of Homeland Security project, and the Federal Bureau of Investigation. With the resulting software, federal forensic examiners performed many high-stakes comparisons of voice recordings. He invented the phonetic refraction method of automatic speaker recognition and used multimodal properties of speech to excel in National Institute of Standards and Technology annual speaker-recognition evaluations. He initiated MIT LL's first Intelligence Advanced Research Projects Activity effort, the Biometrics Exploitation Science and Technology project, resulting in the transition of stand-off biometric technology such as room-audio speaker recognition and long-distance iris recognition. Campbell later led major research innovations in big data analytics, dark web analytics, and counter–human trafficking technology under the Defense Advanced Research Project Agency's XDATA and Memex projects.
Previously, Campbell taught speech processing at Johns Hopkins University (JHU) and served 22 years at the National Security Agency (NSA). At NSA, he and his teammates developed the first DSP-chip software modem and he launched the Biometric Consortium (AFCEA Federal ID Forum and Expo). Jointly with Bell Laboratories, he conducted pioneering work in real-time speech communication over telephone networks, creating Federal Standard 1016 that provided improved voice quality in the government's STU-III secure telephones and the foundations for both digital cellular and voice-over-Internet protocol systems.
Campbell has authored or co-authored more than 120 technical book chapters, proceedings papers, and journal articles, with a total of nearly 8,000 citations. He holds one U.S. patent and led two U.S. Federal Standards, one Federal Information Processing Standard, and one NATO Standardization Agreement. He chairs the advisory committee of JHU's Human Language Technology Center of Excellence, organizes MIT LL's Recent Advances in AI for National Security Workshops, and co-chairs MIT LL's Professional Societies Committee. He was a distinguished lecturer, an editor, and an officer of four positions in the Institute of Electrical and Electronics Engineers (IEEE), and he held two officer positions in the International Speech and Communication Association. Campbell is an IEEE Life Fellow "for leadership in biometrics, speech systems, and government applications." He earned MIT LL's highest award, the Technical Excellence Award, "for his outstanding contributions and leadership in human-language technology, pioneering developments in speech processing, expertise in biometrics, innovations in machine learning, and effective technology transition to government, which have significantly impacted the nation's intelligence, warfighting, and law-enforcement capabilities."
Campbell earned BS, MS, and PhD degrees in electrical engineering from Rensselaer Polytechnic Institute, JHU, and Oklahoma State University (OSU), respectively. For his dissertation, "Features and Measures for Speaker Recognition," he received OSU's Graduate Research Award, recognizing it as the best dissertation across all of OSU's colleges.