I am a computer vision researcher with experience in creating algorithms for point cloud data, RGB and medical images. In 2015, I graduated from a PhD program at the department of electrical and computer engineering from the University of British Columbia (UBC). I have been developing machine learning algorithms at various companies ever since.

My focus is in creating small, easy-to-use and deployable implementations of recent papers in the field of computer vision and machine learning. In my experience, it is extremely difficult to reproduce academic papers in a production environment as most either lack an implementation altogether or they require significant effort for deployment. This is mainly due to the fact that researchers are more focused on publication and are not incentivized to release their code.

The project that has given me the most exposure on Github is pycpd. This is a python implementation of the popular CPD point cloud registration method. Point cloud registration is a ubiquitous term for the process of aligning to point clouds from different sources into a common frame of reference. This is mainly used for aligning laser scans from different angles to produce a more complete representation of an object.

I like to think about the sponsorship program similar to buying coffee for a fellow researcher in exchange for help in your project. I don't want anyone to feel guilty for not supporting me. So this is why I have made my tiers based on coffee prices at my local coffee shop. ☕️