http://www.sigvc.org/bbs/thread-57-1-1.html
1 What is the Goal of Sensory Coding:
介绍了两种sensory coding 的方式 PCA 和 Sparse coding,并比较了他们的优缺点,指出对于大 多数生物信息处理来说多采用sparse coding的方式。
2 Sparse coding with an overcomplete basis set A strategy employed by V1 : (戴纬)
作者examine the neruobiological implications of sparse coding
3 Emergence of Simple-Cell Receptive Field Properties by Learning a Sparse Code for Natural Images: (李海昌)
指出对于自然图像来说,对应于sparse表示的集(也就是字典)为 gabor filter 一类的基。有代码。
4 Regression shrinkage and selection via the lasso (肖鸿飞、康翠翠)
lasso的文章
5 atomic decomposition by basis pursuit (戴玮)
求解sparse的一种方法: linear programming.
6 Non-negative matrix factorization with sparseness constraints (朱飞云)
在non-negative matrix factorization中加入系数的sparse约束,使得分解得到的基更像一个part.
7 Least angle regression
8 sparse_signal_restoration(谷鹄翔)
用iterated soft-thresholding algorithm 求解L1
9 A Fast Iterative Shrinkage-Thresholding Algorithm (谷鹄翔)
求解L1,类似与iterated soft-thresholding algorithm,但速度快
10 Iteratively Re-weighted Least Squares Minimization for Sparse Recovery
用iteratively Re-weighted Least Squares 求解L1约束或者是L_{p}约束
11 K-SVD: An Algorithm for Designing Overcomplete Dictionaries for Sparse Representation (徐元)
K-svd求字典
12 Image denoising via sparse and redundant representations over learned dictionaries (徐元)
sparse 用于denoising
13 Image Super-Resolution as Sparse Representation of Raw Image Patches (刘绍国)
sparse 用于super-resolution
14 Robust Face Recognition via Sparse (徐元)
sparse用于人脸识别
15 Compressed sensing
L0与L1的等价性证明
16 Robust uncertainty principles Exact signal reconstruction from highly incomplete frequency information
L0与L1的等价性证明
17 Enhancing Sparsity by Reweighted l1 Minimization
提出加权L1的方法,在很多情况下比L1要好
18 Supervised Dictionary Learning (汪凌峰)
Close the Loop Joint Blind Image Restoration and Recognition with Sparse Representation Prior:
把人脸数据库中的图像来对去模糊后的人脸图像进行sparse表示,以此来提高对模糊人脸图像的识别
Gabor Feature based Sparse Representation for Face Recognition with Gabor Occlusion Dictionary
Online Detection of Unusual Events in Videos via Dynamic Sparse Coding:
与sparse denosing, superresolution的思路差不多
最近会议上的一些文章:
Online Dictionary Learning for Sparse Coding : 可以解决大量样本学习字典的问题
Optical Flow Estimation Using Learned Sparse Model
Robust Tracking Using Local Sparse Appearance Model and K-Selection
Sparse Approximated Nearest Points for Image Set Classifcation
Sparse Representation for Color Image Restoration:
都有:denoising, demosaicing, inpainting
Supervised Dictionary Learning:
在字典学习中除了重构的目标函数外,还加入了一些判别性的信息。
Learning with Structured Sparsity:
A note on the group lasso and a sparse group:
结构sparse,可能与矩阵的2,1范数相似
其他参考网址:http://dsp.rice.edu/cs