Moinul Islam

My research interests lie in multimodal machine learning, with a focus on integrating visual, textual, and spatio-temporal data for real-world applications. I am particularly interested in foundation models such as LLMs and VLMs and their adaptation for tasks involving environmental intelligence, behaviour-aware systems, and long-form content understanding. My work also explores trustworthy AI through uncertainty modeling and factual consistency, alongside scalable end-to-end ML systems for deploying robust, data-driven solutions in complex real-world settings.