Vyas lab

Our research lies at the intersection of artificial intelligence, materials science, and sustainability. By advancing machine learning methods for vision, language, and multimodal understanding, my group develops intelligent systems that accelerate scientific discovery and address critical challenges in renewable energy, manufacturing, and chemical innovation. Our contributions span from fundamental AI benchmarks and algorithms to domain-specific applications in photovoltaic defect detection, industrial safety, and molecular property prediction. This narrative highlights our progress across three interconnected themes: (1) advancing robustness and efficiency in computer vision and foundation models, (2) enabling AI-driven solutions for photovoltaics and industrial safety, and (3) pioneering multimodal AI for chemistry and critical materials discovery with experimental Validation. The figure highlights the current research directions:

PI, Dr. Shruti Vyas, is an interdisciplinary researcher whose work brings together chemical engineering, computer vision, and artificial intelligence to address challenges in sustainability, energy, and industrial diagnostics. She earned her Ph.D. in Chemical and Biomolecular Engineering from the National University of Singapore and her B.Tech. from IIT (BHU) Varanasi, and also spent three years in industry as a process design engineer at Engineers India Limited (EIL) before moving into academic research. Her postdoctoral work at the University of Central Florida focused on representation learning and visual understanding using deep learning, laying the foundation for her current research. At UCF, she leads a group developing intelligent, data-driven methods for applications such as molecular property prediction, photovoltaic defect detection and recycling, geolocalization, and industrial hazard identification. Guided by the vision of bringing the benefits of AI into experimental science, Dr. Vyas aims to accelerate discovery and create innovative solutions with meaningful real-world impact.