Subjective Evaluation of Texture Synthesis Methods
Dr Martin Kolar will present the Department of Computer Science colloquium with a talk entitled, "Subjective Evaluation of Texture Synthesis Methods".
Example-based Texture Synthesis is a problem in Computer Graphics in which the texture from any input image is extrapolated onto an output object without visible flaws or repetitions. With applications in digital photography, computer games, and manufacturing, the perceived quality of texture synthesis algorithms is an important topic in computer graphics. Numerous algorithms have been proposed to tackle this problem, and the work presented in this talk provides the first rigurous user study to assess their quality. This talk will present the design of experimental materials, the user study, and an analysis of the results. This work has been accepted at Eurographics 2017.
Martin Kolar is a reseracher in Computer Graphics and Vision at Brno Univesity of Technology in Czech Republic. He holds a BSc in Mathematics and Computer Science from King's College London, an MSc in Machine Learning from University College London, and a PhD from the University of Warwick. The topic of his doctoral work is High Quality Texture Synthesis, under the supervision of Prof. Alan Chalmers and Dr. Kurt Debattista.