About
A quarter century of turning signals into understanding
AICIP — the Advanced Imaging and Collaborative Information Processing Lab — is a computer vision and machine learning research group in the Min H. Kao Department of Electrical Engineering and Computer Science at the University of Tennessee, Knoxville.
The lab was built by Dr. Hairong Qi, who joined UT in 1999 and, within a year, was leading a DARPA SensIT project on a question that still defines the group: how do many imperfect observers become one reliable understanding? In 2000 that meant battery-powered acoustic sensors scattered across a desert, sharing mobile software agents instead of raw data. The group was fielding that work at DARPA demonstrations by 2001, and the AICIP name appears in print by 2003.
The name was chosen carefully. Advanced imaging reflects the lab's lineage in imaging systems — Dr. Qi's dissertation work, early thermal-infrared medical imaging, multispectral focal-plane arrays, and the hyperspectral analysis that became a signature strength. Collaborative information processing names the second pillar: distributed inference across networks of resource-constrained sensors, the work recognized in Dr. Qi's IEEE Fellow citation.
Both pillars are still load-bearing. The hyperspectral line now runs on generative models; "collaboration" now means fusing webcams with weather models, radar with cameras, or biometrics with virtual reality — and the sensors have become foundation models. What persists is the lab's habit of treating information as something that must be earned from imperfect, heterogeneous, resource-constrained observation.
The lab today
AICIP is a deliberately compact group — a director, a postdoc, and a focused cohort of PhD students — with an outsized publication record: 200+ papers, two Cambridge University Press textbooks, and recent work at NeurIPS, ICLR, ECCV, WACV, and IJCAI. Current research spans five active areas, from multimodal and self-supervised learning to AI for health and scientific discovery. The lab has graduated more than thirty PhD students, whose careers span Adobe, Apple, Amazon, Ford, Google, national laboratories, and faculty positions.
Within UT, the lab anchors the university's growing AI footprint: Dr. Qi leads the Foundational Artificial Intelligence cluster-hiring initiative and is one of three faculty leaders of ATHENA, UT's $20M NSF-funded automated-materials-discovery testbed.
Timeline
Milestones
A selective history. Every entry below is drawn from primary records — the director's CV, university archives, agency award databases, and the lab's own publications.
-
1999
Dr. Qi joins UT Knoxville
Hairong Qi arrives as an assistant professor in electrical and computer engineering, fresh from a PhD at NC State under Wesley Snyder on a high-resolution, large-area digital imaging system.
-
2000
DARPA SensIT seeds the lab
Qi becomes PI of MU-FASHION, a DARPA SensIT project on multi-resolution data fusion with agent-bearing sensors — the research program the lab is named for.
-
2001
SITEX02 field demonstration
The group fields its mobile-agent collaborative processing work at the DARPA SensIT SITEX02 demo at Twentynine Palms, California, alongside MIT Lincoln Labs, Xerox PARC, BBN, and a dozen universities.
-
2003
The AICIP name in print
The earliest surviving document to use the name — a master’s thesis from the group — describes "The UTK AICIP (Advanced Imaging and Collaborative Information Processing) lab." The same year, the lab’s invited Proceedings of the IEEE article on mobile-agent-based collaborative processing appears.
-
2004
Machine Vision published
Cambridge University Press publishes Machine Vision by Wesley Snyder and Hairong Qi, a widely adopted textbook.
-
2005
NSF CAREER Award
Qi receives the NSF CAREER Award for "Collaborative Signal and Information Processing in Sensor Networks."
-
2007
MVC-NMF
The lab publishes minimum-volume-constrained NMF for hyperspectral unmixing in IEEE TGRS — a paper that becomes a standard of the field.
-
2012
GRSS Highest Impact Paper Award
The IEEE Geoscience and Remote Sensing Society recognizes the 2007 MVC-NMF paper with its Highest Impact Paper Award.
-
2014
Gonzalez Family Professorship
Qi is named the Gonzalez Family Professor in EECS.
-
2017
CVPR spotlight & UTKFace
The conditional adversarial autoencoder for face age progression earns a CVPR spotlight and releases the UTKFace dataset, which becomes a widely used benchmark. Fundamentals of Computer Vision, the lab’s second Cambridge textbook, is published the same year.
-
2018
IEEE Fellow
Qi is elevated to IEEE Fellow "for contributions to collaborative signal processing in sensor networks."
-
2023
Representation learning era
Cross-Scale MAE appears at NeurIPS, anchoring the lab’s self-supervised learning line. Qi is named faculty lead of UT’s Foundational Artificial Intelligence cluster-hiring initiative.
-
2026
ATHENA and a widening frontier
UT wins a $20M NSF award for ATHENA, the university’s AI-driven automated materials discovery testbed, with Qi among its three faculty leaders. Lab papers appear at WACV (FujiView), ICLR (PDT), and ECCV (ExPLoRe).
Within UT EECS
The lab is one of the research groups of the Min H. Kao Department of Electrical Engineering and Computer Science in UT's Tickle College of Engineering, housed in the Min H. Kao Building on the Knoxville campus. Students join through the EECS graduate programs in computer science and in electrical and computer engineering, and the lab regularly involves undergraduate researchers — several of whom have become co-authors, Goldwater scholars, and eventually the lab's own PhD students.
AICIP collaborates widely: with UT colleagues in power systems, robotics, biomedical engineering, and the UT Institute of Agriculture; with Oak Ridge National Laboratory; and with partners from Duke, Northwestern, Johns Hopkins, and industry. Its code and datasets are public at github.com/aicip.
Interested in joining? See how to get involved.