Using a sines + wavelets mixed dictionary for improving matching pursuit-based parametric audio coding

This paper deals with the aplication of adaptive signal models for representing transients and sinusoids at the same stage in a parametric audio coder. To acomplish such goal, we search for sparse aproximations by means of matching pursuit with a mixed dictionary, instead of using two different dictionaries that operate in cascade. In such sense, complex exponentials and wavelets are chosen for modeling the tonal and transient features of anaudio signal, respectively. Experimental results clearly show the advantages of the mixed dictionary over two cascaded dictionaries. The aproach proposed in this paper is sucesfuly applied for parametric audio coding, asuring beter quality than MPEG2/4-AC at 16Kbits/s for most of the tested audio signals.