Homomorphic vector quantisation

A new approach is proposed for vector quantisation in linear predictive speech coding. The problem is formulated as speech model recognition by minimising the Euclidean distance measure of real cepstra of models with unit power transmission. The procedure is robust with respect to quantisation of both the cepstral coefficients and operational results.     

Robust vector quantisation by transmission energy allocation

The authors propose to combat the effects of channel errors on vector-quantised data by optimising the energy allocation for the transmission of codeword symbols. In particular, they demonstrate that the application of transmission energy allocation to the natural binary code obtained from vector quantiser design by the splitting method outperforms established robust techniques such as index assignment, and explicit error control coding of most sensitive bits     

Fast clustering algorithm for vector quantisation

A fast clustering algorithm is presented as an alternative to the K-means algorithm. By encoding training vectors selectively and changing the codebook updating step, the algorithm reduces the computation time. Simulations show that the algorithm outperforms the K-means algorithm in computation time and performance in terms of mean-squared-error     

Efficient algorithm for channel-optimised vector quantisation

An efficient method is presented for increasing the speed of the clustering process for channel-optimised vector quantisation (COVQ), in which the Hamming distances between the indices of codevectors are utilised. Simulation results demonstrate that with a small increase in pre-processing and memory costs, the training time of the new algorithm has been reduced significantly while the encoding quality remains unchanged     

Fast search algorithm for vector quantisation

A tree structure for the fast nearest neighbour search algorithm for vector quantisation is presented. Using the new algorithm a remarkable reduction in the number of multiplications, additions and comparisons is achieved.<>     

New methodology for adaptive vector quantisation

An image sequence coding scheme for low bit-rate video coding is presented. A new methodology is proposed for adaptive vector quantisation (AVQ), where the codebook is updated with new code-vectors. The new code-vectors replace less significant ones in the codebook, based on a novel scoring criterion that utilises a forgetting factor and codebook half-life. The proposed AVQ method gives rise to an additional performance enhancement of approximately 1 dB over frequency scoring techniques     

Fast dead zone lattice vector quantisation

A fast method is proposed to determine the quantisation parameters of an entropy constrained lattice vector, called dead zone lattice vector quantisation (DZLVQ). This scheme offers substantial improvement in terms of visual quality with respect to the well-known SPIHT and JPEG2000 algorithms. The tuning of quantisation parameters in variable rate transform coding scheme may involve a critical increasing of the complexity. It is shown that this problem can be solved by a simple iterative procedure using classical LVQ tuning functions. The algorithm ensures a substantial reduction of the complexity while maintaining the compression performance of DZLVQ.     

Dynamic finite-state vector quantisation of colour images

A dynamic finite-state vector quantisation technique for colour image compression is proposed. The proposed method efficiently exploits both the statistical redundancy between the colour components of a pixel and the high correlations between adjacent pixels. In addition, the image is compressed losslessly as compared with a colour quantised index image. Experimental results show that it can significantly reduce the storage requirement while maintaining excellent image quality.<>     

Adaptive fast encoding algorithm for vector quantisation

A new fast encoding algorithm is developed for vector quantisation (VQ). The proposed algorithm takes the topological structure of the codebook into account and controls the search space adaptively. Simulation results demonstrate that while the performance of the proposed algorithm is the same as that using an exhaustive search, it is more efficient than some of the other related fast VQ encoding algorithms     

Feature-based initialisation for image vector quantisation codebookgeneration

A feature-based initialisation method is developed for codebook design in vector quantisation of images. The proposed algorithm makes use of different features of the image block in training set-to-do clustering for producing the initial codebook for the LBG algorithm. It is shown that the proposed method not only speeds up the LBG process but also enhances the quality of the reproduced images     

Index-compressed vector quantisation based on index mapping

The authors introduce a novel coding technique which significantly improves the performance of the traditional vector quantisation (VQ) schemes at low bit rates. High interblock correlation in natural images results in a high probability that neighbouring image blocks are mapped to small subsets of the VQ codebook, which contains highly correlated codevectors. If, instead of the whole VQ codebook, a small subset is considered for the purpose of encoding neighbouring blocks, it is possible to improve the performance of traditional VQ schemes significantly. The performance improvement obtained with the new method is about 3 dB on average when compared with traditional VQ schemes at low bit rates. The method provides better performance than the JPEG coding standard at low bit rates, and gives comparable results with much less complexity than address VQ     

Pitch predictive vector quantisation of speech waveform

A coding algorithm is presented which combines pitch prediction with low-dimensional vector quantisation to exploit both long- and short-term correlation in the speech waveform at rates of 16 and 9.6 kbit/s. Vector quantisation of the predictor enables the stability of the synthesis filter to be assured, and also allows the use of a minimum residual energy criterion. SNRs of 17-19 dB are achieved at 16 kbit/s and 13-15 dB at 9.6 kbit/s.<>     

Partial-distortion-weighted fuzzy competitive learning algorithmfor vector quantisation

An effective competitive learning algorithm called the partial-distortion-weighted fuzzy competitive learning (PDW-FCL) algorithm is developed for vector quantisation. The PDW-FCL algorithm is able to minimise the likelihood of neuron (codevector) underuse and make good use of every neuron for optimal vector quantiser design     

Pruned tree-structured vector quantisation with dead zones

The authors present pruned tree-structured vector quantisation with dead zones (PTSVQWDZ) which is suitable for transforming image or speech coding. PTSVQWDZ considerably improves the rate-distortion performance with a significantly reduced codebook storage when compared with the conventional PTSVQ     

Predictive pyramid vector quantisation of LSF parameters

The authors propose predictive pyramid vector quantisation (PPVQ) to quantise LSF parameters. For prediction of the LSF parameters, a 4th order moving average (MA) predictor is used. To quantise the prediction error signal effectively, pyramid VQ (PVQ) is considered. The performance of PPVQ has an average spectral distortion (SD) of 1.06 dB at 22bit/frame     

Regularised restoration of vector quantisation compressed images

The authors study the application of image restoration technology in improving the coding performance of a vector quantisation (VQ) image compression codec. Restoration of VQ-compressed images is rarely addressed in the literature, and direct applications of existing restoration techniques are generally inadequate to deal with this problem. A restoration algorithm is proposed, specific to VQ-compressed images, that makes good use of the codebook to derive useful a priori information for restoration. The proposed restoration algorithm is shown to be capable of improving the quality of a VQ-compressed image to a much greater extent, compared with other existing restoration approaches. As no extra information, other than the codebook, is required to carry out the restoration with the proposed algorithm, no transmission overhead is necessary and hence, it can be fully compatible with any VQ codec when used to improve coding performance     

Cache vector quantisation algorithm in video compression

A new vector quantisation (VQ) algorithm based on a strategy similar to the cache in traditional computers is proposed and simulated. The method can significantly reduce the redundancy in VQ indices. According to the simulated results, the bit rate is variable but is much less than that of the primary full search algorithm. For test sequence 'Miss America', the mean square error is approximately 23 and the bit rate is approximately 0.3 bit/pixel.<>     

Systolic encoder for fast waveform vector quantisation

The letter presents a systolic architecture that can efficiently perform the encoding operations of waveform vector quantisation using the squared-error distortion measure. In the system, squared-error cells and comparing cells are used to achieve modularity and concurrency. An index offset approach is employed for reducing I/O pins.     

Fast image colourisation based on twin-codebook vector quantisation

A novel colourisation scheme for grey-level images based on twincodebook vector quantisation (TBVQ) is presented. Different from the traditional vector quantiser with the same codebook in the encoding and decoding processes, the proposed TBVQ adopts two different but related codebooks in the encoder and decoder, respectively. During the colourisation process, the target image is first encoded by the greyscale codebook and then decoded by the related chromatic codebook. Experimental results demonstrate the effectiveness and efficiency of the proposed algorithm.     

Algorithm for hierarchical vector quantisation of video data

A recursive algorithm is presented that is designed to construct quantisation tables and codebooks for the hierarchical vector quantisation of images. The algorithm is computationally inexpensive and yields high quality codebooks