Rate Distortion Behavior of Sparse Sources

被引:32
作者
Weidmann, Claudio [1 ]
Vetterli, Martin [1 ,2 ]
机构
[1] Ecole Polytech Fed Lausanne, Audiovisual Commun Lab, CH-1015 Lausanne, Switzerland
[2] Univ Calif Berkeley, Dept Elect Engn & Comp Sci, Berkeley, CA 94720 USA
关键词
Entropy; memoryless systems; rate distortion theory; sparse signal representations; transform coding; SIDE INFORMATION; NONLINEAR APPROXIMATION; STATIONARY SOURCES; ENTROPY; SIGNALS; BOUNDS;
D O I
10.1109/TIT.2012.2201335
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
The rate distortion behavior of sparse memoryless sources is studied. These serve as models of sparse signal representations and facilitate the performance analysis of "sparsifying" transforms like the wavelet transform and nonlinear approximation schemes. For strictly sparse binary sources with Hamming distortion, is shown to be almost linear. For nonstrictly sparse continuous-valued sources, termed compressible, two measures of compressibility are introduced: incomplete moments and geometric mean. The former lead to low- and high-rate upper bounds on mean squared error, while the latter yields lower and upper bounds on source entropy, thereby characterizing asymptotic behavior. Thus, the notion of compressibility is quantitatively connected with actual lossy compression. These bounding techniques are applied to two source models: Gaussian mixtures and power laws matching the approximately scale-invariant decay of wavelet coefficients. The former are versatile models for sparse data, which in particular allow to bound high-rate compression performance of a scalar mixture compared to a corresponding unmixed transform coding system. Such a comparison is interesting for transforms with known coefficient decay, but unknown coefficient ordering, e.g., when positions of highest-variance coefficients are unknown. The use of these models and results in distributed coding and compressed sensing scenarios are also discussed.
引用
收藏
页码:4969 / 4992
页数:24
相关论文
共 51 条
[1]  
ANDREWS DF, 1974, J ROY STAT SOC B MET, V36, P99
[2]  
Baraniuk R., 2008, IEEE SIGNAL PROCESSI, V25, P21
[3]   Source coding with intermittent and degraded side information at the decoder [J].
Bassi, Francesca ;
Kieffer, Michel ;
Weidmann, Claudio .
2008 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING, VOLS 1-12, 2008, :2941-+
[4]  
Benichou B., 2003, STUDIES COMPUTATIONA, V10, P225
[5]  
Berger T, 1971, Rate Distortion Theory. A Mathematical Basis for Data Compression
[6]   Robust uncertainty principles:: Exact signal reconstruction from highly incomplete frequency information [J].
Candès, EJ ;
Romberg, J ;
Tao, T .
IEEE TRANSACTIONS ON INFORMATION THEORY, 2006, 52 (02) :489-509
[7]   Bounds on the expected length of optimal one-to-one codes [J].
Cheng, Jay ;
Huang, Tien-Ke ;
Weidmann, Claudio .
IEEE TRANSACTIONS ON INFORMATION THEORY, 2007, 53 (05) :1884-1895
[8]   On the importance of combining wavelet-based nonlinear approximation with coding strategies [J].
Cohen, A ;
Daubechies, I ;
Guleryuz, OG ;
Orchard, MT .
IEEE TRANSACTIONS ON INFORMATION THEORY, 2002, 48 (07) :1895-1921
[9]   Nonlinear approximation of random functions [J].
Cohen, A ;
DAles, JP .
SIAM JOURNAL ON APPLIED MATHEMATICS, 1997, 57 (02) :518-540
[10]  
Cover T.M., 2006, ELEMENTS INFORM THEO, V2nd ed