Research

AICIP's research has three focuses, including collaborative information processing with an emphasis in resource-constrained distributed environment, computer vision and machine learning with a focus on unsupervised/self-supervised learning and generative models, and remote sensing with a focus on spectral unmixing. A subset of some very challenging but interesting projects is listed below.

Collaborative Information Processing in Resource-constrained Distributed Environment

  • Autonomous-Driving Vehicles
    • Connected and Autonomous Vehices (CAVs) in the EcoCAR Mobility Challenge (sponsored by DOE)
    • CenterFusion: Center-based Radar and Camera Fusion for 3D Object Detection (GitHub)
    • Investigation of DEtectors, Algorithms, and Systems (IDEAS) to Advance Autonomous Radiological/Nuclear Search (sponsored by DTRA)
  • Information Fusion in Cyber-Physical Systems
    • Robotic Caregiver to Comprehend, Assist, Relieve, and Evaluate for Patients with Alzheimer's Disease (Robotic CARE for AD) (sponsored by NIH, 2022-2027)
    • Secure Constrained Machine Learning for Critical Infrastructure CPS (sponsored by NSF CPS, 2021-2024)
    • Cost-Effective Mastitis Control and Biosecurity for Sustainable Dairy Farming (sponsored by USDA NIFA)
    • Achieving High-Resolution Situational Awareness in Ultra-Wide-Area Cyber-Physical Systems (sponsored by NSF CPS)
  • Collaborative Processing in Sensor Networks
    • Investigation of DEtectors, Algorithms, and Systems for Wearable Intelligent Nuclear Detection (IDEAS for WIND) (sponsored by DHS)
    • Improve Building Efficiency by Equipment Health Monitoring with Virtual Intelligent Sensing (sponsored by DOE)
    • Distributed Solutions to Smart Camera Networks (sponsored by NSF NeTS)
    • Collaborative Signal and Information Processing in Sensor Networks (sponsored by NSF CAREER)
    • An Intelligent Agent-based Model Framework for Optimization and Control of Distributed Sensors (sponsored by ONR - SBIR Phase I and II)
    • MU-FASHION: Multi-Resolution Data Fusion using Agent-Bearing Sensors in Hierarchically Organized Distributed Sensor Networks (sponsored by DARPA SensIT Program)

Computer Vision and Machine Learning

  • Open-set Recognition
    • Representative-Discriminative Learning for Open-set Land Cover Classification of Satellite Imagery (GitHub Repo, ECCV'20)
  • Unsupervised/Self-supervised Learning
    • Cross-Scale MAE: A Tale of Multiscale Exploitation in Remote Sensing (NeurIPS'23)
    • Semantic Segmentation in Aerial Imagery Using Multi-level Contrastive Learning with Local Consistency (WACV'23)
    • Unsupervised and Unregistered Hyperspectral Image Super-Resolution with Mutual Dirichlet-net (TGRS'22)
    • Unsupervised Pansharpening based on Self-Attention Mechanism (TGRS'21)
    • uSDN: Unsupervised Sparse Dirichlet-net for Hyperspectral Image Super-Resolution (GitHub, CVPR'18)
  • Generative Modeling
    • Non-local Representation based Mutual Affine-Transfer Network for Photorealistic Stylization (PAMI'22)
    • UTKFace (GitHub)
    • Face-Aging-CAAE: Age Progression/Regression by Conditional Adversarial Autoencoder (GitHub, CVPR'17)
    • r-BTN: Cross-Domain Face Composite and Synthesis from Limited Facial Patches (AAAI'18)
    • Talking Face Generation by Conditional Recurrent Adversarial Network (GitHub, IJCAI'19)
    • SRNTT: Image Super-Resolution by Neural Texture Transfer (GitHub, CVPR'19)
  • Online Learning
    • Online Knowledge Distillation by Temporal-Spatial Boosting (WACV'22)
    • Online Knowledge Distillation with History-Aware Teachers (IJCNN'22)
    • Derivative Delay Embedding: Online Modeling of Streaming Time Series (GitHub, CIKM'16)

Computer Vision and Machine Learning Applications in Remote Sensing

  • Mutated: Modeling and Understanding using Temporal Analysis of Transient Earth Data (sponsored by IARPA SMART)
    • Cross-Scale MAE: A Tale of Multiscale Exploitation in Remote Sensing (NeurIPS'23)
    • Semantic Segmentation in Aerial Imagery Using Multi-level Contrastive Learning with Local Consistency (WACV'23)
  • LEGO: Large-scale Environment-modeling with Geometric Optimization (sponsored by IARPA CORE3D)
    • Representative-Discriminative Learning for Open-set Land Cover Classification of Satellite Imagery (GitHub Repo, ECCV'20)
  • High Performance Image Processing Algorithms for Current and Future Mastcam Imagers (sponsored by NASA - STTR Phase I and II)
  • Real-Time Smart Tools for Processing Spectroscopy Data (sponsored by NASA - STTR Phase I and II)
  • Spectral Unmixing (sponsored by ONR)
    • uDAS: An untied denoising autoencoder with sparsity for spectral unmixing (GitHub, TGRS'19)
    • MVC-NMF: Unsupervised spectral unmixing by minimum-volume-constrained non-negative matrix factorization (GitHub, TGRS'07)
    • GDME: Unsupervised spectral unmixing by gradient-descent maximum entropy (GitHub, TIP'07)
  • Physically Constrained Transfer Learning through Shared Abundance Space for Hyperspectral Image Classification (TGRS'21)