Who: Dr. Xuming He

 

When: 1:00PM - 2:00PM, Sep. 20, 2016, Tuesday.

 

Where: NICTA Seminar Room. Ground Floor, Building 7A, London Circuit, Canberra, ACT.

 

Abstract:

With increasing demand for efficient image and video analysis, test-time cost of scene parsing becomes critical for many large-scale or time-sensitive vision applications. We propose a dynamic hierarchical model for anytime scene labeling that allows us to achieve flexible trade-offs between efficiency and accuracy in pixel-level prediction.  In particular, our approach incorporates the cost of feature computation and model inference, and optimizes the model performance for any given test-time budget by learning a sequence of image-adaptive hierarchical models.

We formulate this anytime representation learning as a Markov Decision Process with a discrete-continuous state-action space. A high-quality policy of feature and model selection is learned based on an approximate policy iteration method with action proposal mechanism. We demonstrate the advantages of our dynamic non-myopic anytime scene parsing on three semantic segmentation datasets. This is a joint work with Buyu Liu and will appear in ECCV16.

 

Shortbio:

Xuming He is a Senior Researcher at NICTA Canberra Lab and also an adjunct Fellow of Engineering Department at the Australian National University (ANU). He received received the B.Sc. and M.Sc. degrees in electronics engineering from Shanghai Jiao Tong University, China, and Ph.D. degree in computer science from the University of Toronto, Canada. He held a postdoctoral position at the University of California at Los Angeles (UCLA) before joining National ICT Australia (NICTA). His research interests include image segmentation and labeling, visual motion analysis, vision-based navigation, and undirected graphical models.