The image shows the underlying time-frequency characteristics of speech that are exploited by automatic recognition systems.

Artificial Intelligence Technology and Systems

We develop artificial intelligence (AI) algorithms, technologies, and systems for extracting information from multimedia data in adverse conditions. Group members have developed advanced AI-based technologies to achieve world-leading performance in automatic speech,  language, and speaker recognition for automatically identifying the words and language spoken and the speaker. We have successfully applied natural language processing techniques to automatically identify the authors, languages, topics, and entities (such as people, places, organizations, and events) in the content of communications. Our human-network AI technologies automatically extract information from speech, text, image, and video data combined with network communications to infer network and coordinated activities. These advanced detection capabilities are used by the Department of Defense and law enforcement to identify threatening or illicit activity on the surface and dark webs. Recent focus areas for our group include adversarial AI and counter AI to increase analytic robustness and support operations in contested environments. Our group is widely recognized for its strong publications in journals and conferences. We emphasize AI, machine learning, technology transition to government in operational environments, and technology evaluation with operationally relevant metrics and datasets. 

Advancing Our Research


Send us an email to inquire about our career opportunities and to learn more about projects we're working on.

Featured Publications

Poisoning network flow classifiers [e-print]

Jun 2
arXiv:2306.01655v1 [cs.CR]

Improving long-text authorship verification via model selection and data tuning

May 5
Proc. 7th Joint SIGHUM Workshop on Computational Linguistics for Cultural Heritage, Social Sciences, Humanities and Literature, LaTeCH-CLfL2023, 5 May 2023, pp. 28-37.

A generative approach to condition-aware score calibration for speaker verification

Feb 8
IEEE/ACM Trans. Audio, Speech, Language Process., Vol. 31, 2023, pp. 891-901.

Our Staff

View the biographies of members of the Artificial Intelligence Technology and Systems Group.