Browsing by Author "Hussain, Yunus"
Results Per Page
Sort Options
Item Adaptive Block Transofmr Coding of Speech Based on LPC Vector Quantization.(1989) Hussain, Yunus; Farvardin, Nariman; ISRIn this paper we describe an adaptive block transform speech coding system based on vector quantization of LPC parameters. In order to account for the power fluctuations, the speech signal is normalized to have a unit-energy prediction residual The temporal variations in the short-term spectrum, on the other hand, are taken into accour by vector quantizing the UC parameters associated with the vector of speech samples and transmitting the codeword index. For each block based on the codevector associated with the input vector, an optimum bit assignment map is used to quantize the transform coefficients. We consider two types of zero memory quantizers for encoding the transform coefficients, namely the Llyod-Max quantizer and the entropy-coded quantizer. The performance of these schemes is compared with other adaptive transform coding schemes. We show by means of simulations that the system based on entropy-coded quantizer design leads to very high performance and in most cases as much as 5 dB performance improvement in terms of segmental signal-to-noise ratio is observed over the adaptive block transform coding scheme of Noll and Zelinski [1]. The effects of the bit-rate and the size of the codebook on the performance of the systems are also studied in detail.Item Design and Performance Evaluation of a Class of Finite-State Vector Quantizers(1992) Hussain, Yunus; Farvardin, Nariman; ISRThe finite-state vector quantizer (FSVQ), introduced by Foster, Dunham and Gray, is a finite-state machine that can be viewed as a collection of memoryless full-searched vector quantizers, where each input vector is encoded using a vector quantizer associated with the current encoded state; the current state and selected codeword determine the next encoder state. It is generally assumed that the state codebooks are unstructured and have the same cardinality leading to a fixed-rate scheme. In this thesis, we present two variable-rate variations of the FSVQ scheme with the possibility of using structured as well as unstructured state codebooks. In the first scheme, we let the state codebook sizes be different for different states, implying different rate distribution among the states. In the second scheme, in addition to this flexibility, we use pruned tree-structured vector quantizers as the state quantizers, i.e., we let each of the state quantizers be a variable-rate encoder. For encoding sampled speech data, both of these schemes perform significantly better than the fixed-rate FSQV scheme with the second scheme giving the best performance.We also consider that 2-D extension of the above mentioned schemes and describe two bit rate image coding systems based on these schemes. A compression ratio in excess of 26 is achieved for encoding the 512 x 512 version of "Lena" using the schemes employing variable-rate FSVQs.
An implicit assumption made in all the systems mentioned above is that the channel is noiseless. Under noisy channel conditions, all of the above systems suffer from severe performance degradations calling for the need to reformulate the FSVQ design problem taking into account the channel noise. Using some of the developments in joint source-channel trellis coding, we describe two different methods leading to two types of noisy channel FSVQs. We show by means of simulations on the Gauss- Markov source and speech LSP parameters and for a binary symmetric channel that both schemes are fairly robust to channel noise. In particular, for speech LSP parameters, the proposed noisy channel FSVQs lead to a saving of 1.5-4 bits/frame over the channel-optimized vector quantizer depending on the level of noise in the channel.
Item Finite-State Vector Quantization for Noisy Channels(1992) Hussain, Yunus; Farvardin, Nariman; ISRUnder noiseless channel conditions and for sources with memory, finite-state vector quantizers (FSVQs) exhibit improvements over memoryless vector quantizers. It is shown, however, that in the presence of channel noise, the performance of the FSVQ degrades significantly. This suggests that for noisy channels, the FSVQ design problem needs to be reformulated by taking into account the channel noise. Using some of the developments in joint source-channel trellis coding, we describe two different methods leading to two types of noisy channel FSVQs. We show by means of simulations on the Gauss-Markov source and speech LSP parameters and for a binary symmetric channel that both schemes are fairly robust to channel noise. For the Gauss-Markov source, the proposed noisy channel FSVQs perform at least as well as or better than the channel-optimized VQ, while for speech LSP parameters, they lead to saving of 1.5-4 bits/frame over the channel-optimized VQ depending on the level of noise in the channel.Item Variable-Rate Finite-State Vector Quantization(1990) Hussain, Yunus; Farvardin, Nariman; ISRA finite-State vector quantizer is a finite-state machine that can be viewed as a collection of memoryless full-searched vector quantizers, where each input vector is encoded using a vector quantizer associated with the current encoder state; the current state and codeword selected determine the next encoder state [1]. In [1], the state codebooks are unstructured. In addition, it is assumed that all the state codebooks have the same cardinality leading to a fixed-rate system. In this paper, we present two variable-rate variations of the system in [1]. In the first system we let the state codebook sizes be different for different states. In the second system along with the flexibility of having different codebook sizes for different states, we use pruned tree-structured vector quantizers [2] as the state quantizers. For encoding samples speech data, both of these schemes perform significantly better than the scheme in [1]. The second system gives the best performance of all. Performance improvements of up to 4.25 dB at the rate of 5/8 bits per sample are obtained.Item Variable-Rate Finite-State Vector Quantization and Applications to Speech and Image Coding(1991) Hussain, Yunus; Farvardin, Nariman; ISRA finite-state vector quantizer is a finite-state machine that can be viewed as a collection of memoryless full-searched vector quantizers, where each input vector is encoded using a vector quantizer associated with the current encoder state; the current state and selected codeword determine the next encoder state. It is generally assumed that the state codebooks are unstructured and have the same cardinality leading to a fixed-rate scheme [1]. In this paper, we present two variable-rate variations of the scheme in [1] with the possibility of using structured as well as unstructured state codebooks. In the first scheme, we let the state codebook sizes be different for different states, implying different rate distribution among the states. In the second scheme, in addition tot his flexibility, we use pruned tree- structured vector quantizers as the state quantizers, i.e., we let each of the state quantizers be a variable-rate encoder. For encoding sampled speech data, both of these schemes perform significantly better than the fixed-rate scheme of [1]. The second scheme gives the best performance of all; performance improvements of up to 4.25 dB at the rate of 0.5 bits/sample are obtained over the scheme in [1].We also consider the 2-D extension of the above mentioned schemes and describe two low bit rate image coding systems based on these schemes. The first system subtracts the mean from each input block and then encodes the mean-subtracted block by means of the 2-D versions of fixed- rate and variable-rate finite-state vector quantizer; the block- mean is separately encoded in an efficient manner by exploiting the high correlation present in the means of adjacent blocks. In the second system, a prediction is made on each pixel using a 5th-order predictor and the residual is again encoded using the 2-D versions of the fixed-rate and variable-rate finite-state vector quantizer. At a bit rate of 0.3 bits per pixel, a peak signal-to-noise ratio in excess of 31 dB is achieved for encoding the 512 x 512 version of "Lena" using the schemes employing variable-rate finite-state vector quantizers.