The main focus of my research is on developing new unsupervised and supervised machine learning techniques for segmentation and visualization of 3D images. Visualizing 3D images (e.g. MRI, CT-Scan data) properly is still a major challenge due to the complex nature of the raw data. My target is to combine the new techniques to extract important information from these images and use intuitive user interfaces for easy visualization with automated color assignment, occlusion handling, segmentation etc.
I was an intern at Microsoft Research Asia
from October 2013 to January 2014. There I have worked on developing a
skeleton based performance evaluation framework for Microsoft Kinect,
which can be used for physical rehabilitation and sports training. This
project also inrporporated the principles of visualization to provide
the user an easily interpretable performance feedback. A demo will be posted on this website soon.
My Masters thesis was on efficient Content-Based Image Retrieval (CBIR).
Conferences and Workshops