I am a Senior Lecturer in the Department of CSE at Ahsanullah University of Science and Technology (AUST). My research focuses on developing intelligent, trustworthy, and human-centered AI systems that address real-world challenges through interdisciplinary approaches. My primary research interests include Natural Language Processing, AI Alignment, Human-AI Interaction, Vision-Language Models, and Deep Learning. I am particularly interested in advancing large language models and multimodal AI, with an emphasis on aligning AI systems with human values, improving human–AI collaboration, and developing robust solutions for language and vision understanding.
I welcome opportunities to collaborate with academic and industry partners on innovative research and applications in computational linguistics and artificial intelligence. If you are interested in exploring potential collaborations, please feel free to contact me via email.
") does not match the recommended repository name for your site ("").
", so that your site can be accessed directly at "http://".
However, if the current repository name is intended, you can ignore this message by removing "{% include widgets/debug_repo_name.html %}" in index.html.
",
which does not match the baseurl ("") configured in _config.yml.
baseurl in _config.yml to "".

Most. Sharmin Sultana Samu, Md. Rakibul Islam, Md. Zahid Hossain, Md. Kamrozzaman Bhuiyan, Farhad Uz Zaman
28th International Conference on Computer and Information Technology (ICCIT 2025)
TL;DR: We present the first systematic benchmark for Bengali deepfake audio detection, showing that fine-tuned deep learning models significantly outperform zero-shot approaches, with ResNet18 achieving state-of-the-art performance on the BanglaFake dataset.
Most. Sharmin Sultana Samu, Md. Rakibul Islam, Md. Zahid Hossain, Md. Kamrozzaman Bhuiyan, Farhad Uz Zaman
28th International Conference on Computer and Information Technology (ICCIT 2025)
TL;DR: We present the first systematic benchmark for Bengali deepfake audio detection, showing that fine-tuned deep learning models significantly outperform zero-shot approaches, with ResNet18 achieving state-of-the-art performance on the BanglaFake dataset.

Md. Zahid Hossain, Farhad Uz Zaman, Md. Rakibul Islam
26th International Conference on Computer and Information Technology (ICCIT 2023)
TL;DR: We detect AI-generated images using optimized CNNs, achieving 96.31% accuracy, with Grad-CAM providing insights into model decisions.
Md. Zahid Hossain, Farhad Uz Zaman, Md. Rakibul Islam
26th International Conference on Computer and Information Technology (ICCIT 2023)
TL;DR: We detect AI-generated images using optimized CNNs, achieving 96.31% accuracy, with Grad-CAM providing insights into model decisions.