Superfast Wavelet Transform Using QTT Approximation. I: Haar Wavelets
Boris N. Khoromskij and Sentao Miao
Contact the author: Please use for correspondence this email.
Submission date: 14. Nov. 2013 (revised version: November 2013)
published in: Computational methods in applied mathematics, 14 (2014) 4, p. 537-553
DOI number (of the published article): 10.1515/cmam-2014-0016
MSC-Numbers: 65F30, 65F50, 65N35, 65F10
Keywords and phrases: tensor-structured methods, fast wavelet transform, canonical tensor decomposition, quantized tensor approximation (QTT), data compression, multilevel methods
Download full preprint: PDF (271 kB)
We propose a superfast discrete Haar wavelet transform (SFHWT) as well as its inverse, using the QTT representation for the Haar transform matrices and input-output vectors. Though the Haar matrix itself does not have a low QTT-rank approximation, we show that factor matrices used at each step of the traditional multilevel Haar wavelet transform algorithm have explicit QTT representations of low rank. The SFHWT applies to a vector representing a signal sampled on a uniform grid of size N = 2d. We develop two algorithms which roughly require square logarithmic time complexity with respect to the grid size, O(log 2N), hence outperforming the traditional fast Haar wavelet transform (FHWT) of linear complexity, O(N). Our approach also applies to the FHWT inverse as well as to the multidimensional wavelet transform. Numerical experiments demonstrate that the SFHWT algorithm is robust in keeping low rank of the resulting output vector and it outperforms the traditional FHWT for grid size larger than a certain value depending on the spacial dimension.