发表日期：2017-09-11 点击击数: 382
报告题目：Regularization: the Key in Machine Learning based Image Recognition
Images contain rich information. However, a specific recognition task needs only a small portion of it embedded and hidden in the rich image information. Even in a narrow domain of image recognition, face recognition for example, different information is required for face detection, emotion recognition, gender recognition and identity recognition. This brings the difficulty for us to design a robust recognition system by the limited human knowledge due to the complex characteristics of image and large variations of images taken under different conditions. Machine learning from the data is a solution to find right information from the high dimensional image for specific recognition tasks. It is thus not a surprise that learning-based approaches emerge in various applications. The powerfulness of machine learning was already proven 30 years ago in the boom of neural networks but its successful application to the real world is just in recent years. This is because the machine learning alone can only solve problems in the training data but the recognition system is designed for the unknown data outside of the training set. This gap can be bridged by regularization: human knowledge guidance or interference to the machine learning. This speech will analyze these concepts and ideas from subspace approaches of dimensionality reduction, sparse coding, to deep CNN neural networks.
Xudong Jiang received B.Eng. and M.Eng. from the University of Electronic Science and Technology of China and the Ph.D. from Helmut Schmidt University, Hamburg, Germany. He received two Science and Technology Awards from the Ministry for Electronic Industry of China. From 1998 to 2004, he was with the Institute for Infocomm Research, A*STAR, Singapore, as a lead Scientist, and the Head of the Biometrics Laboratory, where he developed a system that achieved the highest efficiency and the second highest accuracy at the International Fingerprint Verification Competition in 2000. He joined Nanyang Technological University (NTU), Singapore, as a Faculty Member in 2004, and served as the Director of the Centre for Information Security from 2005 to 2011. He is currently a Tenured Associate Professor with the School of EEE, NTU. He holds 7 patents and has authored over 150 papers with over 50 journal papers, including 5 TPAMI papers, 9 TIP papers and 13 Pattern Recognition papers. His research interests include image processing, pattern recognition, computer vision, machine learning, and biometrics. He is IFS TC member of the IEEE Signal Processing Society, and serves as Associate Editor of IEEE Transactions on Image Processing, IEEE Signal Processing Letters and IET Biometrics.