FujiView
Will you see Mount Fuji tomorrow? FujiView fuses live webcam imagery with weather data to forecast scenic visibility — published at WACV 2026.
Advanced Imaging & Collaborative Information Processing
From many views,
one understanding.
AICIP is the computer vision and machine learning lab of Dr. Hairong Qi at the University of Tennessee, Knoxville. For twenty-five years the lab has studied how reliable understanding emerges from imperfect, distributed observation — from DARPA-era sensor networks to today's multimodal AI.
One scene, five realities — webcam views from FujiView, the lab's scenic-visibility forecasting system published at WACV 2026.
25+
years of research
200+
technical publications
31
PhD graduates
10+
federal sponsors
Featured projects
What the lab is building now
All projects MUSICARE-VR
Music intervention in virtual reality that adapts to a person with dementia in real time — with reinforcement learning and AI agents driving socially assistive robots.
Representation Learning
From Cross-Scale MAE to ExPLoRe — a sustained line of work on learning visual representations without labels, published at NeurIPS 2023 and ECCV 2026.
ATHENA
A $20M NSF-funded testbed making UT a hub in the national network of AI-powered cloud laboratories for accelerated materials discovery — with Dr. Qi as one of three faculty leaders.
Research areas
The lab's intellectual range
About our researchMultimodal & Self-Supervised Learning
Fusing heterogeneous data — images, weather, text, sensors — and learning strong representations without labels.
Computer Vision & Machine Learning
Recognition, detection, and generative modeling — the lab's core discipline for two decades.
Remote Sensing & Hyperspectral Imaging
Understanding Earth from above — hyperspectral unmixing, satellite change detection, and foundation models for overhead imagery.
AI for Health & Assistive Technology
Adaptive systems that support people — VR music intervention for dementia, socially assistive robots, and environmental health sensing.
Learning Dynamics & Reinforcement Learning
Treating neural network training as a dynamical system — Koopman operator theory for faster training and more stable policy optimization.
AI for Science, Agriculture & Mobility
Putting the lab's methods to work — automated materials discovery, animal-welfare monitoring, and vehicle autonomy.
Selected publications
Recent & landmark work
All publicationsFujiView: Multimodal Late-Fusion for Predicting Scenic Visibility
IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) 2026
ExPLoRe: Expert Patch-Level Loss Routing for Multi-Objective Masked Image Modeling
European Conference on Computer Vision (ECCV) 2026
KIPPO: Koopman-Inspired Proximal Policy Optimization
International Joint Conference on Artificial Intelligence (IJCAI) 2025
Cross-Scale MAE: A Tale of Multiscale Exploitation in Remote Sensing
Advances in Neural Information Processing Systems (NeurIPS) 2023
CenterFusion: Center-based Radar and Camera Fusion for 3D Object Detection
IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) 2021
People
A compact lab with range
Everyone, including alumni
Bryce Bible
PhD Student, Computer Science
Multimodal learning & fusion · Computer vision · VR & assistive AI for health
Weisheng Tang
Postdoctoral Research Associate
Training dynamics · Remote sensing · Multimodal biomedical data
Landon Harris
PhD Student, Electrical Engineering & Computer Science
Reinforcement learning · Learning dynamics & diffusion · Sensor fusion
Andrei Cozma
PhD Student, Electrical Engineering & Computer Science
Reinforcement learning · Koopman operator methods · Self-supervised learning
Shah Md Nehal Hasnaeen
PhD Student, Bredesen Center
Machine learning · Multimodal fusion · RF & electromagnetic sensing
Hameed Ijadunola
Graduate Student
Representation learning · Vision–language models
News
Latest from the lab
All newsThree AICIP PhDs completed in 2026
Konstantinos Georgiou, Maofeng Tang, and Fanqi Wang all defended and deposited their dissertations this year — spanning masked representation learning and the dynamics of neural training.
UT wins $20M NSF award for ATHENA — with Dr. Qi among its three faculty leaders
The NSF selected UT as one of twenty national hubs in its Programmable Cloud Laboratories network. ATHENA will apply AI to automated materials discovery, with AICIP providing the machine learning core.
ExPLoRe accepted to ECCV 2026
The lab's latest self-supervised learning work routes per-patch training signal through Soft-MoE dispatch weights — closing out a three-paper arc from Cross-Scale MAE through MEDiC.
Sponsors & partners
NSF · DARPA · IARPA · ONR · NASA · NIH · DOE · DHS · DTRA · USDA
Federal agencies that have supported AICIP research across 25 years. See the funding record.
Work with us
The lab looks for students who want to do careful, consequential research in computer vision and machine learning.
How to join