Having worked in Machine Learning, Image Processing and parallel programming I have a good grasp on implementing algorithms on both CPU's and GPU's. However, personally my favourite research area is deep neural networks and computer vision.
Languages I have picked up over the years.
Tools I use for a living.
The goal of this project was to smoothly morph one face to another face using the thin plate spline, radial basis function. Morphing combines both shape and color of reference images. It is pretty similar to what you see in movies when a face morphs into another character.
Check it out
Searching for exotic particles like the Higgs-Boson in a high energy particle collider is essentially a complex signal-versus-background classification problem. Hence in this project we, a team of two, used Machine learning in python with pandas, scikit-learn and tensorflow to produce meaningful results. Cclassification algorithms like deep neural networks, logistic regression and decision trees implemented in python on the UCI Machine Learning ‘HIGGS Data Set’ were used to predict whether an output signal from a collider, is a Higgs-Boson Particle signal or simply background noise. The results showed that deep neural network of 9 layers worked best with the accuracy of 75%.
Check it out