Projects & Grants
[2024] CIVIC-PG Track A: Co-Created Pathways: Establishing Community and First Responder Collaboration for Equitable Evacuation and Emergency Sheltering During Extreme Floods
Funding Agency: National Science Foundation (NSF)
Award Number: 2431452
Date: 10/01/2024-03/31/2025.
[2024] Future Scenarios of Phytoplankton and Harmful Algal Bloom Dynamics in the Mississippi-Atchafalaya River System: Leveraging PACE, Optics-Inspired Deep Learning and Coupled Modeling
Funding Agency: National Aeronautics and Space Administration (NASA)
Date: 07/01/2024 - 06/30/2027.
[2024] Federated Deep Learners for Medical Image Analysis and Treatment Guidance
Funding Agency: Louisiana Board of Regents (BOR)
Date: 06/01/2024 - 06/30/2027
[2023] New Faculty Startup Fund
Funding Agency: University of Louisiana at Lafayette (ULL)
Date: 08/01/2023 - 07/31/2025
Research Interests & Directions
Advanced Machine Learning Frameworks for Next-Generation Wireless Communications and Internet-of-Things Intelligence
Artificial Intelligence has recently emerged as an important tool for data processing in modern wireless communications system and IoT applications. To leverage the power of machine learning, we aim to develop efficient AI approaches for wireless communications and IoT intelligence, with topics including:
Generative AI for Semantic Communications
Federated Learning for Distributed IoT systems
Mixture-of-Experts for Edge Computing
Graph Neural Network for Structured Data
Data-Driven Radiomap Estimation, Spectrum Management and Resource Allocation
Providing rich information in propagation behavior and spectrum occupancy, radiomaps could provide detailed spatial information on radio PSD distribution, which consequently benefit applications of resource allocation and spectrum management. We aim to develop novel physics-inspired learning machines for radiomap estimation, together with its applications in network optimization.
Learning and Signal Processing over High-Dimensional Graphs
To uncover the underlying multilateral and multil-level structures for complex systems and big data, we extend the traditional GSP to high-dimensional graph domains, including hypergraph signal processing (HGSP) and multilayer graph signal processing (M-GSP). We focus on the development of theoretical fundamentals and practical applications for HD graph neural networks and graph signal processing.
Physics-Inspired Machine Learning for Remote Sensing and Geoscience Applications
To leverage the physics models in machine learning to address the limited quantity and unsatisfying quality of data in realistic applications, we aim to develop physics-inspired machine learning by integrating physical principles and remote sensing data, with broad applications in carbon, phytoplankton, and nutrions in coastal waters.
AI-Empowered Internet-of-Thing Intelligence for Flood Monitoring and Disaster Control
To enable the emerging machine learning technologies and IoT intelligence in fast and robust emergency response, we aim to develop novel algorithms for flooding monitoring, human behavior analysis, decision making, and UAV controling in Water Resource Engineering. The innovations focus on advanced methods of computer vision, reinforcement learning, large foundation models and game theory.
Acknowledgement to our Funding Sponsor
Office: Madison Hall 248H
Tel: (337) 482-1300
Email: songyang.zhang@louisiana.edu
Department of Electrical & Computer Engineering
Madison Hall 146b
131 Rex St., Lafayette, LA 70504