About Me

I am an Undergraduate AI Researcher pursuing my B.Tech in Computer Science & Engineering (Artificial Intelligence and Data Science) at the Manipal Institute of Technology, Sikkim.

My research centers on building efficient, scalable deep learning systems for high-degradation and low-resource environments. I specialize in applying novel state-space models (SSMs), self-supervised pretraining, and uncertainty-aware learning across diverse domains ranging from OCR to biomedical signal processing and medical imaging to complex computer vision and document analysis. Driven by end-to-end ML workflows, I actively open-source my architectures and am fiercely preparing for masters admissions to advance foundational AI representations.

Core Research Themes

• State Space Models & Sequence Modeling: Optimizing 2D and 3D Mamba architectures for spatial, temporal, and volumetric modeling.
• Uncertainty Quantification: Integrating conformal prediction and fuzzy logic for provable reliability in medical and satellite imagery.
• Efficient Architectures: Developing lightweight, high-throughput architectures via geometric residual learning and DDL overwrites for edge-device deployment.
• Medical Imaging & Remote Sensing: Developing robust frameworks resistant to distribution shifts and data leakage. Utilizing light-weight & efficient architectures for remote sensing and earth observations.
• Computer Vision & OCR: Generative Adversarial Networks, Diffusion Models, Tesseract, Vision Transformers, and Deep Delta Learning applications for image restorations, degraded document analysis & scene text recognition.