讲座题目：《Sparse Coding and Dictionary Learning for Signal and Image Processing》
主 讲 人：段晔博士
美国密苏里大学工程学院都市安全研究中心主任， 计算机系计算机图形与图像理解实验室主任， 副教授、吉林大学唐敖庆讲座教授。
1: Introduction of Regularization, Convexity, L1-minimization, L0-Norm.
2: Motivation and the Sparse Coding Algorithm.
3: Greedy Pursuit Algorithms: Matching Pursuit (MP), Orthogonal Matching Pursuit (OMP), Least Squares OMP (LS-OMP), Weak Matching Pursuit (WMP), and Thresholding algorithm.
4: Dictionary Learning: the K-SVD Algorithm.
5: Image Denoising and Image Inpainting with a learnt Dictionary.
6: Image Compression and Compressive Sensing.
In the past few years significant progress has been made in Sparse Representation: an efficient representations of data as a (often linear) combination of a few typical patterns (atoms) learned from the data itself. Sparse representation has led to state-of-the-art results in many signal and image processing and analysis tasks. This course will describe effective algorithms for learning such collections of atoms (usually called dictionaries or codebooks), and how to compute such sparse representations with high accuracy. We will discuss both the theoretical foundations of Sparse Representation as well as the state-of-the-art in applying these techniques in areas such as medical imaging, computer animation, computer graphics, computer vision and virtual reality, etc. We will explore current research issues and will cover in depth the associated computational and numerical techniques. This course should be appropriate for graduate students in all areas as well as advanced undergraduate students.