Research Projects


Surgical Skills Assessment (2014-present)

pic

The project focuses on developing an automated framework for assessment of the expertise level of surgeons using the OSATS (Objective Structured Assessment of Technical Skills) criteria. We capture RGB-D and accelerometer data of surgical residents and nurse practitioners with differing skill levels. The work uses different computer vision and machine learning techniques to classify each trial as a beginner, intermediate or an expert. Some of the techniques being used includes HMM, BoW, A-BoW, Motion Texture(MT), Sequential Motion Texture (SMT) and frequency analysis.

Following are links to project pages for different publications under this project:

MICCAI 2015 – Surgical Skills Assessment Using Frequency Analysis


Previous:

Sonar Tracking of Indus River Dolphins (2011-2012)

This project was done as the final year project of my undergraduate at LUMS. Indus River Dolphins are one of the few river dolphins in the world. However, the rapid decrease in their population over the years has brought these dolphins on the brink of extinction. The main aim of the project was to develop a acoustic-based system for the survey of the endangered Indus river dolphins so that the movements of the species can be tracked in order to prevent the dolphins from going into regions which are not favorable for their survival. The full report for this project can be found on the link below.

Sonar Tracking of Indus River Dolphins

Machine Learning for Color Classification and Image Segmentation (Summer 2011)

This project was done in a summer internship in my junior year of undergrad in 2011 under the supervision of Dr. Sean Mcloone in National University of Ireland, Maynooth. This project addressed the issue of color classification with specific application to Robo-Cup Soccer League. Machine Learning tools like Linear Discriminant Analysis and Support Vector Machines were used though some other techniques involving different combinations of RGB and the principal components were also analysed. Please click on the link below to check the report for the project.

SIAR Report


Ball and Plate system (Summer 2010)

This project was done in a summer internship in my sophomore year of undergrad in 2011 under the supervision of Dr.Ismail Lazoglu in KoC University, Istanbul. The project focused on developing a ball and plate system from scratch; the mechanical structure and electrical system. The built system was then used to test different algorithms. The work from this internship was published in the International Conference on Electrical and Control Engineering (ICECE), Sep 2011. Please click on the link to check out the paper.

Polar and Polygon Path Traversal of a Ball and Plate System