Data Science and Machine Learning Group

Home Research Publications News

News


Congratulations to Yue Dong and Philippe Paradis, respectively a current and recent MSc student in Mathematics at U Ottawa. They won 3rd place in the Data Day 3.0: Graduate Student Poster Competition held Tuesday March 29th at Carleton University. The next day, Wednesday March 30th, they won 2nd place in the Graduate Student Poster Competition at the Faculty of Engineering and Computer Science of the University of Ottawa. Their winning project used Gram Matrices of Feature Maps in Convolution Neural Networks (CNN) as a representation of the artistic style of an input template and then made use of this encoded “style” to produce a rendition in that “artist’s style” of an arbitrary input image. Their work built on work of Gatys et al (2015) together with modifications of the usual CNN architecture for improved efficiency. Details can be found in their poster.