Brought up in a pure environment surrounded by mountains and woods. I deem myself as a lucky person born with natural gifts Some of them are "ordinary", showed obviously on math, music and memory at a quite early age and continued to serve as parts of my distinctiveness as I grew up some of them are "extraordinary ", for example, my original interest in philosophy. It is hard to believe to some that what I have speculated about at the age of five coincided with some of the thoughts in Jean-Paul Sartre's "L'Etre at le Neant".
It still remains a mystery to me how I left childhood dream of become an astronomer and chose electronic engineering as my college major instead. Not knowing much about circuits as about stars at that time. I seemed to be simply allured by the idea of designing a guidance system for the space shuttle. Five years of inquiry into the electronic world proved that my choice was right. I was completely infatuated with such a field-so profound and vigorous that one is only limited by his/her imagination.
Anxious to find out what "electronics" really means. I began doing laboratory research as early as from the third year. First at the Microwave Lab, doing circuits simulation and user interface design. Generally speaking. Works during that period were fragmentary , but they have given me sufficient experience on C. C++, Windows programming and interface design, which laid a sold foundation for my future study.
Later, to explore something new, I applied to join the Multimedia Lab, an "interactive "place sponsored by Prof Xing Li, I became familiar with UNIX system and did X Windows interface design for some time Probably because some interesting points in my design, I was offered an opportunity by Prof Li to join the "wavelet research group". Intending to develop wavelet applications .
That was a turning point in my work. Wavelet is a brand-new area with modern mathematical background, free space of imagination and, a little but philosophical color that totally interested me I carried out a experiment of applying Embedded. Zero tree Wavelet (EZW) algorithm to speech coding EZW algorithm was originally developed in image coding, aiming at the multi-resolution transmission of pictures. Its efficiency greatly depends on the amount of zerotrees the source produced . By properly quantized, speech samples can be expected to generate a considerable number of zerotrees due to its spectrum characteristics . In my experiment, different kinds of speech sources have been tested and an average compression ratio of 8 I was achieved, with no perceptible distortions Compared with previous coding methods, the result is encouraging, for the benefits are gained through a rather concise procedure doing encoding and decoding is as schematically as doing a cross -word puzzle Moreover. What I have done is actually superficial. There still exist at least several means to improve it, for example, the application of non-even quantizer, wavelet packet, the definition of a higher dimensional structure like "zeroforest ". the selection of an appropriate wavelet base, etc.. The improvement of the algorithm is included in my present research.