Dr. Tianfu (Matt) Wu
NC State University
Modern technological advances produce data at breathtaking scales and complexities such as the images and videos on the web. Such big data require highly expressive models for their representation, understanding and prediction. To fit such models to the big data, it is essential to develop practical learning methods and fast inferential algorithms. My research has been focused on learning expressive hierarchical models and fast inference algorithms with homogeneous representation and architecture to tackle the underlying complexities in such heterogeneous big data from statistical perspectives. In this talk, I will first show our development of a restricted Visual turing test system. Then, I will use online object tracking as a running task to explain my methods of teaching a computer to learn interpretable models and fast inference algorithm in a cooperative manner. To address the limited bandwidth in the current visual Turing test, I will present my on-going and future work on life-long communicative learning based on situated dialogue which integrates the deep perception of visual content and the perception of “dark matter” including human’s beliefs, intents, goals and even values.
Tianfu (Matt) Wu joined NC State in August 2016 as a Chancellor’s Faculty Excellence Program cluster hire in Visual Narrative. Wu, an assistant professor in the Department of Electrical and Computer Engineering, researches explainable and improvable visual Turing test and robot autonomy through lifelong communicative learning. To accomplish his research goals, he pursues a unified framework for machines to ALTER (Ask, Learn, Test, Explain and Refine) recursively in a principled way. Recently, his work was focused on the following: statistical learning of large scale and highly expressive hierarchical and compositional models from visual big data (images and videos); statistical inference by learning near-optimal cost-sensitive decision policies; statistical theory of performance guaranteed learning algorithm and optimally scheduled inference procedure; and statistical framework of visual Turing test and lifelong communicative learning.
Wu received his Ph.D. in statistics from the University of California, Los Angeles (UCLA). He was a postdoctoral researcher in the Department of Statistics at UCLA. Prior to joining the NC State faculty, he was a research assistant professor of statistics in the Department of Statistics at UCLA.
View Presentation Slides