Title：Identifying and modeling what to share for computer vision and machine learning
Speaker：Dr. Wei-Lun (Harry) Chao
Time：106年01月04日(三)10:10~11:10am
Place：交通大學工程三館015室(EC015)
Abstract
The availability of large-scale labeled data is one of the key factors that drive recent advance in computer vision and machine learning. There are many application scenarios, however, where collecting and labeling a large set of representative instances can be laboriously difficult and costly, such as recognition in the wild, structured prediction, and personalized systems. It is thus important to make efficient and effective use of data.
In this talk, I will introduce the concept of sharing---across categories, models, tasks, modalities, or even datasets---which can potentially take the best advantage of existing data. I will start with a global overview, and then focus on how to identify and model “what to share” for vision and learning. Finally, I will show examples where I apply such a concept, including zero-shot learning for visual recognition, video summarization, and probabilistic inference.
Biosketch
Wei-Lun (Harry) Chao is a 4th-year Ph.D. student in Computer Science at University of Southern California, where he is advised by Professor Fei Sha. He received a M.S. and a B.S. in Communication Engineering at National Taiwan University and National Chiao Tung University, respectively. His research interests are in computer vision and machine learning. He is particularly interested in transfer learning, structure learning, and semantic modeling. His personal website is http://www-scf.usc.edu/~weilunc/index.html
