Skip to content
View TapasKumarDutta1's full-sized avatar
3 followers · 1 following

Achievements

Achievement: YOLOAchievement: QuickdrawAchievement: Pull Shark

Achievements

Achievement: YOLOAchievement: QuickdrawAchievement: Pull Shark

Block or report TapasKumarDutta1

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Please don't include any personal information such as legal names or email addresses. Maximum 100 characters, markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
TapasKumarDutta1/README.md

👋 Hi, I'm Tapas Kumar Dutta

Computer Vision and Machine Learning Researcher

I am a Master's student at the University of Surrey currently working on my thesis at the SketchX Lab under the guidance of Subhadeep Koley and Professor Yi-Zhe Song. My research centers on Deep Learning, Medical Imaging, and Foundation Models—developing AI-driven diagnostic tools, advanced X-ray classification systems, and innovative approaches for sketch understanding.

I have actively contributed to multiple international research teams, collaborating with esteemed groups such as SWLEOC, Bagci Lab, and INRIA, STARS Team. My research experience spans collaborations across the US, UK, France, Germany, Japan, and India.

🔬 Research & Publications

  • SAM-Mamba
    Mamba Guided SAM Architecture for Generalized Zero-Shot Polyp Segmentation (WACV 2025)
  • CDANet
    Dual Attention-Based CNN for Brain Tumor Classification (ICIP 2022)
  • GT-Net
    Global Transformer Network for Multiclass Brain Tumor Classification
  • ARM-Net
    Attention-Guided Residual Multiscale CNN for Brain Tumor Classification

🏆 Competitions & Awards

  • Google Landmark Recognition 2020: Achieved Top 11%
  • Data Science Bowl 2019: Ranked within Top 14%

📌 Notable Research Projects

  • Japanese Manga Translation:
    Developed an integrated pipeline using DETR for text detection, TrOCR for recognition, and Transformer models for translation.
  • Anomaly Detection:
    Implemented an unsupervised Student-Teacher Model paired with HistAuGAN to achieve robust anomaly detection.

📫 Connect with Me

I am enthusiastic about potential research collaborations and opportunities. Please feel free to reach out if you would like to discuss research ideas or opportunities.

Popular repositories Loading

  1. Manga_translation Manga_translation Public

    An AI-driven project for manga translation, employing DETR for object detection, Trocr for text recognition on Manga109 dataset, and a Transformer architecture to translate Japanese text to English.

    Jupyter Notebook 6

  2. ARM-Net ARM-Net Public

    This study introduces ARMNet, a sophisticated attention-based model designed to enhance the efficacy of early-stage brain tumor detection through advanced feature extraction methodologies.

    Python 3 1

  3. SAM_Mamba_2025 SAM_Mamba_2025 Public

    [WACV 2025] SAM-Mamba: Mamba Guided SAM Architecture for Generalized Zero-Shot Polyp Segmentation

    3

  4. anomaly-detection anomaly-detection Public

    Unsupervised Anomaly Detection Utilizing a Teacher-Student Model Enhanced by Generative Adversarial Networks.

    Jupyter Notebook 2

  5. HuBMAP HuBMAP Public

    Blood Vessel Instance Segmentation in Human Kidney Histology Slides: A Repository for Comprehensive Whole-Slide Image Analysis

    Jupyter Notebook 1 1

  6. IEEE-CIS-Fraud IEEE-CIS-Fraud Public

    Credit card fraud classification using handcrafted features, feature extraction algorithms and ensemblin.

    Jupyter Notebook