No abstract

This paper presents a new LPC parameter quantization method -SubBand Synthesized LPC Vector Quantization (SBS-LPC-VQ) [9]. In the subband synthesis process, the relationships between subband spectra and whole band LPC spectrum are established and thus the vector quantized subband LPC parameters are mapped to the whole band LPC parameters. This new SBS-LPC-VQ method overcomes high complexity problem of vector quantization of LPC parameters and isolates the distortion within each subband during the VQ process. It also provides the flexibility of assigning the bits to be used for each subband, choosing the order for LPC filter and determining the number of bands for the subband classification. The Critical band Weighting Spectral Distortion Measure (CW-SDM), which is perceptuallymotivated objective measure by using a critical band weighting function, is used for measuring the distortion of quantized LPC spectrum. Using this kind of distortion measure, the experiments show that the SBS-LPC-VQ has obtained 24 bitdframe for coding whole 16th order LPC parameters with about 1 dB average spectral distortion. For comparison, the results by conventional Spectral Distortion Measure (SDM) are also presented in the paper.

This paper proposes an efficient method for estimating frame energy of speech from Enhanced Variable Rate Coder (EVRC) bitstream for network-based speech processing applications in Transcoder Free Operation (TrFO) environments, where speech signals are represented as speech coding parameters. A frame of speech energy is decomposed into the energy of excitation and vocal tract filter, and the frame energy estimation method is derived for each component. Among many parameters of EVRC bitstream, the fixed codebook gain and adaptive codebook gain are used for the estimation of excitation energy, and Line Spectrum Pair (LSP) information is used to estimate the energy of vocal tract filter. Experimental results demonstrated the novelty of the proposed method. The correlation coefficient between the actual and estimated frame energy can be maintained at a value of 0.994 with just 5% multiplicative operations of full decoding.