Md. Tahmid Islam Tomal

Computer Vision & Machine Learning

My research interests center on vision-language models and representation learning, with a focus on designing efficient, reliable, and secure architectures that adapt to new concepts and environments. I aim to understand how learned representations shape model behavior when evidence is limited, ambiguous, or differs from training data. My research has explored these questions through few-shot and incremental learning, while my engineering work has exposed practical challenges in retrieval, context management, and multimodal reasoning. I hold a B.Sc. in Computer Science and Engineering from BUET, where my undergraduate thesis explored attention mechanisms and YOLO models for road crack detection. I currently work as a Machine Learning Engineer at Panjeree Publications and seeking PhD opportunities in computer vision and machine learning.