2012 Poster Sessions : Supervised Earth Mover’s Distance Learning and Its Computer Vision Applications

Student Name : Fan Wang
Advisor : Leo Guibas
Research Areas: Graphics/HCI
Earth Mover’s Distance (EMD) is an intuitive and natural distance metric for comparing two histograms or probability distributions. We propose to jointly optimize the ground distance matrix and the EMD flow-network based on partial ordering of histogram distances in an optimization framework. Two applications in computer vision are used to demonstrate the effectiveness of the algorithm: firstly, we apply the optimized EMD value to face verification, and achieve state-of-the-art performance on public face data sets; secondly, we use the learned EMD flow-network to analyze the internal structure of a set of faces, and consistent paths that demonstrate intuitive transitions on certain facial attributes are found.

Fan Wang is a 3rd year PhD student in Department of Electrical Engineering. She is now advised by Professor Leonidas Guibas in Geometric Computing group, and works on computer vision and machine learning methods. Her current projects include face verification/recognition, image attribute analysis, etc. Before joining Stanford, Fan obtained her B.Eng and M.S. degree both from Tsinghua University in China.