Adaptive synthesis of a wavelet transform using fast neural network
Jan Stolarek · Bulletin of the Polish Academy of Sciences Technical Sciences · 2011
Adaptive synthesis of a wavelet transform using fast neural network This paper introduces a new method for an adaptive synthesis of a wavelet transform using a fast neural network with a topology based on the lattice structure. The lattice structure and the orthogonal lattice structure are presented and their properties are discussed. A novel method for unsupervised training of the neural network is introduced. The proposed approach is tested by synthesizing new wavelets with an expected energy distribution between low- and high-pass filters. Energy compaction of the proposed method and Daubechies wavelets is compared. Tests are performed using sound and image signals.