Seminar: Regularized Deep Learning with Data Geometry and Filter Structures
Assistant Research Professor, Duke University
Monday, February 3, 2020
10:00am - 11:00am
655 McBryde Hall
The central problem of deep learning is how to generalize well from training data to unseen data. One such solution is to regularize deep learning with priors encoded into models. In this talk we will discuss various techniques we recently developed in regularizing deep learning with data geometry, such as low-rank subspace, or structures over convolutional filters. We will show in this way critical yet missing properties can be brought into deep learning, including rotation equivariance, stochasticity, and domain invariance. We will present several real-world examples in cross-spectral face recognition, image hashing, image segmentation, and 3D object detection.
Qiang Qiu received his Bachelor's degree with first class honors in Computer Science in 2001, and his Master's degree in Computer Science in 2002, from National University of Singapore. He received his Ph.D. degree in Computer Science in 2013 from University of Maryland, College Park. During 2002-2007, he was a Senior Research Engineer at Institute for Infocomm Research, Singapore. He is currently an Assistant Research Professor with the Department of Electrical and Computer Engineering, Duke University. His research interests include computer vision and machine learning, specifically on deep learning, image understanding, representation learning.