I’m adding some of the projects close to my heart. For all the project descriptions please visit my LinkedIn profile or Github.

 

Heart Rate Estimation

 

The aim of the heart rate estimation project was to estimate the heart rate of a person through a facial video, in order to invent a non-invasive technique for measuring the heart rate. The algorithm reads the pixel color change in the forehead region of the person and finds an average RGB value which is then mapped to the frequency domain through Fourier Transform to get the power spectrum. The algorithm correctness was verified by taking different videos (before and after a gym session) and estimating the heart rate first through a heart rate monitoring device and then running the algorithm on the video to see the prediction from the heart rate algorithm. It was observed there was +/- 10% variance.

 

Technologies and Programming Languages:
OpenCV 3.0 and Python 3.

 

United Nations Millenium Development Goals:

 

The aim of this project was to solve the problem outlined in the driven-data competition on Millennium Development Goals. We had to sort through the different goals from a time frame of 1972 to 2007 and predict the values for 2008 and 2012. We achieved good accuracy and were placed 112 in the all time list.

 

Technologies and Programming Languages:
Python 3, Machine Learning – Linear Regression, Support Vector Regression – using Apache Spark’s ML library. on a Hadoop Cluster with 5 nodes.