Evolving next generation signal compression and reconstruction transforms via genetic algorithms

Ongoing research has established a new methodology for using genetic algorithms [2] to evolve forward and inverse transforms that significantly reduce quantization error in reconstructed signals and images. The approach promises to revolutionize the signal and image processing field, producing both higher quality images and higher compression ratios than is currently possible with wavelet-based techniques. Categories and Subject Descriptors J.2 [Physical Sciences and Engineering]: Engineering. General Terms Algorithms, Performance, Experimentation. Keywords Genetic algorithms, wavelets, optimization, evolved transforms.

• Type: Conference
• Year: 2005
• Where: GECCO
• Authors: Frank W. Moore, Patrick Marshall
• Keywords: GECCO 2005, General Terms Algorithms, Genetic Algorithms, Keywords Genetic Algorithms