Research · Current direction
M.Tech Dissertation Phase ILearning to ask
better questions.
My research interests sit across artificial intelligence, machine learning, deep learning, computer vision, and medical imaging. I am currently working on Phase I of my M.Tech dissertation, developing the research foundation before full-scale experimentation.
Current academic work
M.Tech Dissertation
Phase I
This phase is about building a defensible foundation: understanding existing literature, identifying a focused problem, choosing sound methods, and defining how the work will be evaluated.
Research scope is currently being refined. A formal title and detailed findings will be shared when the work is mature enough to represent accurately.Literature review
Studying relevant work, comparing methods, and mapping the limitations that leave room for a meaningful research question.
Problem formulation
Narrowing the scope, defining objectives, and turning a broad area of interest into a question that can be tested properly.
Methodology
Planning datasets, preprocessing, baselines, candidate models, evaluation metrics, and a reproducible experimental workflow.
Initial experiments
Implementing the first baselines and using their results to refine assumptions before deeper experimentation in the next phase.
Areas of interest
Where I'm focusing
my attention.
Machine & deep learning
Model design, representation learning, evaluation, optimization, and the practical path from an experiment to a dependable intelligent system.
Computer vision
Teaching machines to interpret visual information through classification, detection, segmentation, and robust feature learning.
Medical imaging
Exploring how learning-based vision methods can support the analysis of medical images while respecting reliability, interpretability, and clinical context.
Applied AI research
Research that joins technical rigor with usable software—clear baselines, reproducible experiments, honest limitations, and real-world relevance.
Research principles
Useful results require honest methods.
I'm early in this journey, but the standard I want to develop is clear: understand the problem, establish fair baselines, document decisions, measure what matters, and communicate uncertainty.
Understand prior work before claiming novelty.
Use reproducible experiments and meaningful comparisons.
Show limitations alongside promising results.
Research meets implementation