Home

About me:

Welcome!

My name is Patrick, I am a Principal Data Scientist at Microsoft, leading various Computer Vision / Machine Learning projects with teams of up to 10 people. This includes our public Computer Vision best-practices repository which recently reached 2,500 stars on Github.

Before joining Microsoft, I worked at a startup called Tanden
t Vision Science. Until 2010, I was a PhD student and a member of the VGG team at the University of Oxford. My supervisors were Prof. Andrew Zisserman, Dr. Mark Everingham and Prof. Sir Michael Brady.

Prior to my PhD studies, I spent an amazing year in Japan working for Asahi Kasei on visual drowsiness recognition. Furthermore, I was jointly funded by a Fulbright and German State scholarship to study for a year at the University of Massachusetts, Amherst. I received my Master's degree (German: Diplom) in Computer Science & Computer Engineering from the University of Mannheim, Germany.


My work:

My main software development interests are in the Deep Learning, Computer Vision, Machine Learning, Natural Language Processing, and Medical Image Processing fields.

During my time at Microsoft I have been working on a variety of Computer Vision and Natural Language Processing projects, including object detection for Liebherr or Coca-Cola, image similarity ranking for Brillio, or domain classifiers and named entity taggers for Bing. While at Tandent, I have been working on detecting and understanding illumination in images. In my free time, I wrote and published a Windows Phone app called "Piano Teacher".

My PhD was about the recognition of sign language in video sequences. The aim was to develop a classifier which can automatically learn and recognize Sign Language signs in continuous signing sequences, i.e. without separations between signs. The New Scientist, Focus online (in German) and the University of Oxford wrote an article about my PhD work.

See the "Past projects" section for more information.

Updates:


2017-05: Slides from my talk at the Open Data Science conference can be downloaded from here.
2017-05: Slides from my talk at the Microsoft AI immersion workshop can be downloaded from here.
2017-06:
Slides from my talk at the O'Reilly Artificial Intelligence conference can be downloaded from here.

Last updated: February 2018