Generalized Boundary Adaptation Rule for minimizingr - th power law distortion in case of high resolution quantization
نویسندگان
چکیده
A new generalized unsupervised competitive learning rule is introduced for adaptive scalar quan-tization. The rule, called generalized Boundary Adaptation Rule (BAR r), minimizes r-th power law distortion D r in the high resolution case. It is shown by simulations that a fast version of BAR r outperforms generalized Lloyd I in minimizing D 1 (mean absolute error) and D 2 (mean squared error) distortion with substantially less iterations. In addition, since BAR r does not require generalized centroid estimation, as in Lloyd I, it is much simpler to implement.
منابع مشابه
Generalized boundary adaptation rule for minimizing rth power law distortion in high resolution quantization
A new generalized unsupervised competitive learning rule is introduced for adaptive scalar quantization. The rule, called generalized Boundary Adaptation Rule (BARr), minimizes r-th power law distortion Dr in the high resolution case. It is shown by simulations that a fast version of BARr outperforms generalized Lloyd I in minimizing D1 (mean absolute error) and D2 (mean squared error) distorti...
متن کاملA Ziv - Zakai - Rényi Lower Bound on Distortion at High Resolution
We follow a method introduced by Ziv and Zakai for finding ‘informational’ lower bounds on delay constrained joint source-channel coding. Their method uses the data processing theorem for generalized measures of information. We introduce the use of Rényi’s information of order α in their framework, and use high-resolution approximations to find its rate distortion function for a source that pos...
متن کاملSemilogarithmic Nonuniform Vector Quantization of Two-Dimensional Laplacean Source for Small Variance Dynamics
In this paper high dynamic range nonuniform two-dimensional vector quantization model for Laplacean source was provided. Semilogarithmic A-law compression characteristic was used as radial scalar compression characteristic of two-dimensional vector quantization. Optimal number value of concentric quantization domains (amplitude levels) is expressed in the function of parameter A. Exact distorti...
متن کاملExtension of two-stage vector quantization-lattice vector quantization
This paper is the extension of two-stage vector quantization–(spherical) lattice vector quantization (VQ–(S)LVQ) recently introduced by Pan and Fischer [1]. First, according to high resolution quantization theory, generalized vector quantization–lattice vector quantization (G-VQ–LVQ) is formulated in order to release the constraint of the spherical boundary for the second-stage lattice vector q...
متن کاملA generalized interpolative VQ method for jointly optimal quantization and interpolation of images
In this paper we discuss the problem of reconstruction of a high resolution image from a lower resolution image by a jointly optimum interpolative vector quantization method. The interpolative vector quantizer maps quantized low dimensional 2x2 image blocks to higher dimensional 4x4 blocks by a table lookup method. As a special case of generalized vector quantization (GVQ), a jointly optimal qu...
متن کامل