"Towards Understanding Online Visual Media" by Dr. Gang Hua

Date: November 20, 2009
Time: 1:00 pm – 2:00 pm

Where: BSB 3.02.02 (Loeffler Room)

"Towards Understanding Online Visual Media"
by Dr. Gang Hua (Nokia Research Center)


The proliferation of internet and visual sensors have created tremendous amount of online visual media. To assist the users to more conveniently access and share these visual media, it is an emerging need to intelligently analyze, understand and therefore annotate the content of them. It is not a trivial problem due to the unconstrained setting, essentially we are facing all the possible challenges a computer vision system can expect.

In this talk, I will present how four elements, namely feature, learning, context and users, can be synergistically leveraged to approach to understanding massive real-world visual media. In particular, I will present our latest work on learning computational efficient and yet compact local image descriptors. In addition, I will also present one case study: contextual face recognition for online photo annotation, to demonstrate how different levels of context, e.g., spatial, temporal, and social level context, can be leveraged to improve computer vision algorithms. Last but not least, I will briefly discuss the importance to have human in the loop, especially in the online scenario, where community efforts can be fostered to facilitate our final goal, i.e., to understand massive real-world visual media. If time permitted, in the endof the talk, I will also present one slide summary of several other works related to this topic.


Dr. Gang Hua is a Researcher in Nokia Research Center, Hollywood. Before that, he was a Scientist in Microsoft Live labs Research from 2006 to2009. He obtained his Ph.D. degree in Electrical and Computer Engineering from Northwestern University in 2006. His current research interests include computer vision, pattern recognition, machine learning, pervasive media computing, human computer interaction, large scale multimedia information retrieval, mining andsearch. He is a member of both IEEE and ACM. As of August, 2009, he holds 1 US patent and has 16 more patents pending.