Current View
Citation link:
Плешкова, Снежана Георгиева; Pleshkova, Snezhana Georgieva; Момчеджиков, Михаил Благоев рец.; Momchedzhikov, Mihail Blagoev rew.
Невронна мрежа за търсене в celp кодовата книга чрез оценка на линейните спектрални двойки (LSP)
Other Titles:
Neural network for celp code book search estimation of line spectral pair (LSP)
Date of Issue:
Is Part of:
PROCEEDINGS OF THE TECHNICAL UNIVERSITY - SOFIA, 55, 2005 - Anniversary “Sixty Years Technical University - Sofia”, pp. 195-202
TU - Sofia
0374-342Х [issn]
невронни мрежи - приложение
The search process in CELP code book is a sequence of error estimation of each code book vector, if this candidate-vector is used in vector quantization of original speech signal. The result of this sequence process is the choice of a code book vector, for which the book search is the most time consuming operation in CELP coding. Therefore, in this article it is proposed to use a neural network for the realization of CELP code book search process and error estimation with line spectral pairs. In the beginning of this article it is made a choice of more suitable type of the neural network for the code book search. It is proposed a neural network structure for implementation in the standard CELP algorithm for speech coding. Final it is made a MATLAB simulation for testing of the proposed neural network CELP code book search method with using the line spectral pairs.
Рец. проф. д-р Михаил Момчеджиков. - Със сх., диагр.