Souvik Ghosh ☕️
Souvik Ghosh

Lecturer

About Me

I am Souvik Ghosh, a prospective graduate student for Fall 2026, with a strong background in computer science and engineering. I currently serve as a Lecturer in the Department of Computer Science and Engineering at BRAC University and completed my undergraduate studies in CSE at BUET. My research interests span bioinformatics—particularly proteomics and structural bioinformatics. I am driven by the pursuit of developing novel computational and machine learning approaches to address complex biological and real-world problems, aiming to contribute to both scientific knowledge and practical applications.

Download CV
Interests
  • Computational Biology
  • Proteomics
  • Structural bioinformatics
Education
  • BSc. in Computer Science and Engineering

    Bangladesh University of Engineering and Technology

  • HSC

    Dhaka College

  • SSC

    Dhanmondi Govt. Boys' High School

📚 My Research

I am Souvik Ghosh, a Lecturer in the Department of Computer Science and Engineering at BRAC University. I completed my undergraduate studies in CSE at BUET. I am passionate about applying machine learning and deep learning to solve impactful problems in computational biology and beyond. My undergraduate thesis focuses on post-translational modification prediction, where I developed ResLysEmbed, a ResNet-based framework that integrates sequence data and protein language model embeddings to achieve state-of-the-art results in succinylation site prediction.

Beyond bioinformatics, I have explored deepfake detection using generative models and advanced vision architectures, as well as multimodal learning for cancer prognosis prediction, combining genomic and clinical data. My broader research interests span LLMs, generative models, and multimodal AI, with a strong motivation to bridge computational methods and biological insights.

I aspire to continue advancing AI-driven solutions in healthcare and scientific discovery while collaborating across disciplines to push research frontiers. 😃

Featured Publications
Recent Publications
(2025). DFCON: Attention-Driven Supervised Contrastive Learning for Robust Deepfake Detection. arXiv.
Recent News

🥈 1st Runner Up - IEEE SP Cup 2025

Our team ‘Straw Hats’ secured the 1st runner-up position at IEEE SP Cup 2025 with our work on DFCon for deepfake detection.

🎓 Thesis Defense Completed

Successfully defended my undergraduate thesis on ResLysEmbed, a ResNet-based framework for lysine succinylation site prediction.