Tapas Kumar Dutta

I am a Master's student at the University of Surrey, where I have been working on my thesis at the SketchX Lab under the supervision of Subhadeep Koley and Professor Yi-Zhe Song.

I'm also a Data Scientist at National Health Service, SWLEOC, while collaborating with Bagci Lab and INRIA, STARS Team on various research projects.

Prior to this, I worked as a Deep Learning Engineer at 2SigmaSchool, specializing in Langchain, Retrieval-Augmented Generation, PDF Parsing;

Alongside, I have held research and engineering positions at LearnOpenCV, BitsCrunch, VIVEN, Malaviya National Institute of Technology Jaipur and Indian Institute of Technology, Hyderabad.

Tapas Kumar Dutta

Email  /  LinkedIn  /  GitHub  /  Kaggle /  Google Scholar

Research Interests

I specialize in Deep Learning, focusing on Medical Imaging, Foundation Models, and Generative AI. My work includes AI-driven diagnostics, X-ray analysis, sketch understanding, GAN-based augmentation, and AI applications in NFT fraud detection and valuation.

Research
2025
SketchFusion: Learning Universal Sketch Features through Fusing Foundation Models New!
S.Koley, T. K. Dutta, A. Sain, P.N. Chowdhury, A.K. Bhunia, Y-Z. Song
In Proc. IEEE / CVF Computer Vision and Pattern Recognition Conference ( CVPR ), 2025
[PDF] / [BibTeX] / [arXiv] / [Code] /
SAM-Mamba: Mamba Guided SAM Architecture for Generalized Zero-Shot Polyp Segmentation New!
T. K. Dutta, S. Majhi, D. R. Nayak, D. Jha
In Proc. IEEE/CVF Winter Conference on Applications of Computer Vision ( WACV ), 2025
[PDF] / [BibTeX] / [arXiv] / [Code] /
2024
GT-Net: global transformer network for multiclass brain tumor classification using MR images
T. K. Dutta, D. R. Nayak, R.M. Pachori
In Biomedical Engineering Letters, 2024
[PDF] / [BibTeX] / [arXiv] / [Code] /
ARM-Net: Attention-guided residual multiscale CNN for multiclass brain tumor classification using MR images
T. K. Dutta, D. R. Nayak, Y.D. Zhang
In Biomedical Signal Processing and Control, 2024
[PDF] / [BibTeX] / [arXiv] / [Code] /
2022
CDANet: Channel split dual attention based CNN for brain tumor classification in MR images
T. K. Dutta, D. R. Nayak
In 2022 IEEE international conference on image processing ICIP
[PDF] / [BibTeX] / [arXiv] / [Code] /