ICASSP 83: Proceedings, April 14, 15, and 16, 1983, Sheraton-Boston Hotel, Boston, Massachusetts, Volume 3 |
From inside the book
Results 1-3 of 91
Page 971
... obtained by adding the output of several elements located on the edge of the array because these elements are closest to the source of the flow noise . Note that in figure 2 , this sample is obtained from an accelerometer which is a ...
... obtained by adding the output of several elements located on the edge of the array because these elements are closest to the source of the flow noise . Note that in figure 2 , this sample is obtained from an accelerometer which is a ...
Page 1027
... obtained at the expense of an increase in memory occupation . A comparison in terms of memory requirements among these two alternatives of Vector Quantization application is shown in Fig . 1 , where also the case of no quantization is ...
... obtained at the expense of an increase in memory occupation . A comparison in terms of memory requirements among these two alternatives of Vector Quantization application is shown in Fig . 1 , where also the case of no quantization is ...
Page 1043
... obtained using the non - probabilistic approach , and the best network obtained using the probabilistic technique described in this paper ( the latter for the two sizes of training set ) . Table I also shows the overall percentage ...
... obtained using the non - probabilistic approach , and the best network obtained using the probabilistic technique described in this paper ( the latter for the two sizes of training set ) . Table I also shows the overall percentage ...
Common terms and phrases
Acoust adaptive algorithm all-pole amplitude analysis applied approximation ARMA array ASSP autocorrelation average band bandlimited bandwidth baseband beamformer bit rate block cepstrum channel codebook coder coding coefficients components compression computed convolution correlation covariance DCT-I decoder deconvolution defined detection distortion DPCM dynamic dynamic time warping entropy equations error rate filter bank finite frame frequency function ICASSP IEEE IEEE Trans implementation input inverse isolated word iterative length Levinson Recursion linear prediction matrix measure method obtained operations optimal output parameters pattern performance phase pitch problem Proc procedure ratio recursive reference reflection coefficients residual samples segment seismic sequence signal processing sinusoids speaker spectral estimation spectrum speech recognition speech signal sub-band synthesis technique template tion utterance values variable variance vector quantization vocabulary vowel waveform wavelet white noise word recognition zero