Abstract: Recently, vector quantization has become noted as a highly efficient coding method of image and voice data. So far, many of the highly efficient coding problems, or service coding problems, have been studied separately from channel coding problems. This paper reconsiders vector quantization jointly optimizing source coding and channel coding, and proposes a new vector quantizer for noisy channels. Vector quantizers for binary symmetric channels are designed for memoryless Gaussian source, Gauss-Markov source and the real images, and are compared with the conventional vector quantizer which does not take account of channel errors. As a result, it is shown that the performance of the proposed vector quantizer is improved without adding the redundancy for error correction, and the improvement of the performance of the proposed vector quantizer for noisy channels over the conventional vector quantizer becomes significant for highly correlated sources and longer block length.

Abstract: A fundamental computational task that arises in several areas of signal processing is pattern matching, where a given test pattern is compared with a large set of stored templates, to find the best match that minimizes a given measure of dissimilarity. Three different geometrically-oriented methods are proposed for substantially reducing the computational complexity of the search process by reducing the number of multiplies in exchange for additional low complexity operations and, in two of the methods, additional memory for storing precomputed tables.

Abstract: A vector quantizer maps a k -dimensional vector into one of a finite set of output vectors or "points". Although certain lattices have been shown to have desirable properties for vector quantization applications, there are as yet no algorithms available in the quantization literature for building quantizers based on these lattices. An algorithm for designing vector quantizers based on the root lattices A_{n}, D_{n} , and E_{n} and their duals is presented. Also, a coding scheme that has general applicability to all vector quantizers is presented. A four-dimensional uniform vector quantizer is used to encode Laplacian and gamma-distributed sources at entropy rates of one and two bits/sample and is demonstrated to achieve performance that compares favorably with the rate distortion bound and other scalar and vector quantizers. Finally, an application using uniform four- and eight-dimensional vector quantizers for encoding the discrete cosine transform coefficients of an image at 0.5 bit/pel is presented, which visibly illustrates the performance advantage of vector quantization over scalar quantization.

Abstract: : A new approach to isolated-word speech recognition using vector quantization (VQ) is examined. in this approach, words are recognized by means of sequences of VQ code books called multi-section code books. A separate multi-section code book is designed for each word in the recognition vocabulary by dividing the word into equal-length sections and designing a standard VQ code book for each section. Unknown words are classified by dividing them into corresponding sections, encoding them with the multi-section code books, and finding the multi-section code book that yields the smallest average distortion. For speaker-independent recognition of a 20-word vocabulary containing the digits, this approach achieves 95% recognition accuracy for the full vocabulary and 99% for the digits, in both causes with approximately 90% fewer distortion computations than typical dynamic-time-warping approaches. In addition, the approach achieves greater than 99% accuracy for speaker-dependent recognition of the digits with only 1 distortion computation per input frame per vocabulary word. The approach is described, detailed experimental results are presented and discussed, and computational requirements are analyzed. (Author)

Abstract: This paper introduces a Hierarchical Vector Quantization (HVQ) scheme that can operate on "supervectors" of dimensionality in the hundreds of samples. HVQ is based on a tree-structured decomposition of the original super-vector into a large number of low dimensional vectors. The supervector is partitioned into subvectors, the subvectors into minivectors and so on. The "glue" that links subvectors at one level to the parent vector at the next higher level is a feature vector that characterizes the correlation pattern of the parent vector and controls the quantization of lower level feature vectors and ultimately of the final descendant data vectors. Each component of a feature vector is a scalar parameter that partially describes a corresponding subvector. The paper presents a three level HVQ for which the feature vectors are based on subvector energies. Gain normalization and dynamic codebook allocation are used in coding both feature vectors and the final data subvectors. Simulation results demonstrate the effectiveness of HVQ for speech waveform coding at 9.6 and 16 Kb/s.

Abstract: A necessary and sufficient condition is presented for the normalized asymptotic r th power distortion of a mismatched multidimensional quantizer to converge to a certain optimum constant known as Bennett's integral. The dimensionality of our results is allowed to approach infinity in order to make some universal source coding comparisons to quantization theory.

Abstract: The encoding of colored pictures in components has attracted a lot of attention. In this paper, for their efficient transmission in the 2-3 bit/pel range, the direct encoding of the red ( R ), green ( G ), and blue ( B ) primaries is investigated. Differential PCM with vector quantization is used and the encoding efficiencies are compared to the similar result obtained using the properly converted luminance ( Y ) and chrominance ( I,Q ) component signals. A new quantization scheme is proposed for RGB encoding to improve the picture quality. As a result, it is shown that RGB encoding can attain a comparable or slightly higher encoding efficiency as compared to YIQ encoding, and a higher resolution can be obtained for strong color edges defined by the primary colors, but the resulting picture quality is much noisier; thus, it is better suited for such applications in which the edges of primary colors play important roles, such as in graphics materials. An adaptive vector quantization is introduced in order to apply the proposed RGB encoding scheme to a wider range of pictures. The quantity called activity is introduced, considering the form of the predictor used. The optimum adaptive scheme is derived and its effectiveness is verified through simulations.

Abstract: This paper concerns a new Quantization Scheme for efficient encoding of waveforms below or about one bit per sample. This technique is then applied to the encoding of baseband residual signals to demonstrate the feasibility of (baseband) Residual Excited LPC at 2400 hit/sec. The technique called "Spherical Vector Quantization" is described in which a block of n consecutive samples is quantized as a vector. The magnitude of the vector is transmitted independently of the vector's orientation. This vector's orientation is vector quantized using a codebook which can be seen as representing a set of N points on a unit hypersphere. The cases for blocks of n = 8 and 24 are discussed which make use of results by Conway and Sloane on regular point lattices. For n = 8, the algorithm is detailed which solves both the problem of finding the closest point on the hypersphere and the problem of determining the index of that point. Experiments involving SVQ in coding the speech baseband residual are described which show in particular that subband coding does not contribute any quality improvement when SVQ is used.

Abstract: Locally optimum vector quantizer (VQ) designs are presented for memoryless Gaussian, gamma, and Laplacian sources For Gaussian sources, low (2-6) dimensional vector quantization provides relatively little improvement in mean-squared error (MSE) compared to the minimum mean-squared error (MMSE) scalar quantizer For Laplacian or gamma sources, however, significant improvement in MSE is available with vector quantization The Laplacian and gamma 6 bit, sixdimensional vector quantizers achieve, respectively, improvements of 2 and 45 dB over the corresponding scalar MMSE quantizer distortions

Abstract: Vector quantization (VQ) has made it possible to utilize perceptually meaningful techniques for direct space-domain image coding. A simple 2 or 3 way classified codebook approach [2,3] allocates the perceptually important edges with more resolution than the easily encoded monotone regions of an image. In this paper, we introduce a major extension of the classification approach to include edge orientation and location, thereby exploiting an important feature of the human visual mechanism. In particular, each 4 × 4 block of pixels is classified into one of 31 classes for the case of 16 dimensional VQ. The encoding and codebook design complexity is significantly reduced, allowing us to use large codebooks designed from a large database of training images. We present images encoded at 0.7 and 0.8 bits per pixel using this scheme with 16- dimensional vectors. Only a small fraction of one bit per pixel is needed to code the monotone regions of an image; the rest of the bitrate is used to achieve a high level of edge integrity.

Abstract: Vector quantization (VQ) is examined as a technique to enhance performance in subband coding of speech at 9.6 kb/s. The set of short-term subband power levels is vector quantized, providing low-rate side information to control the coding of the subband signals. Each subband signal is then vector quantized with variable size codebooks that are dynamically assigned by the quantized side information. Two versions are described, a 7-band coder and a 14-band coder. Simulation results demonstrate that vector quantization offers a distinct perceptual improvement compared with scalar quantization of the same subband signals and side information for the same total bit rate.

Abstract: Recent work at AT&T Bell Laboratories has shown how the theories of Vector Quantization (VQ) and Hidden Markov Modeling (HMM) can be applied to the recognition of isolated word vocabularies. The initial experiments with an HMM word recognizer were restricted to a vocabulary of 10 digits. For this simple vocabulary with dialed-up telephone recordings, we found that a high-performance, speaker-independent word recognizer could be implemented, and that the performance was, for the most part, insensitive to parameters of both the HMM and the VQ. In this paper we extend our investigations of the HMM recognizer to the recognition of isolated words from a medium-size vocabulary (129 words), as used in the AT&T Bell Laboratories airlines reservation and information system. For this moderately complex word vocabulary, we have found that recognition accuracy is indeed a function of the HMM parameters (i.e., the number of states in the model and the number of symbols per state). We have also found that a VQ that includes energy information gives better performance than a conventional VQ of the same size (i.e., same number of code-book entries).

Abstract: A proposed class of vector quantizers with memory, called Predictive Vector Quantizers (PVQ), exhibits good performance (as measured by mean-squared error distortion) for the encoding of speech signals at 16 kbits/sec. Two methods for designing PVQ's are described, and their performance is compared to that of memoryless VQ's by computer simulation. These results indicate that, for comparable performance, the PVQ has a smaller block size and computational load than the memoryless VQ. In addition, preliminary simulations indicate that PVQ's are less speaker dependent than memoryless vector quantizers.

Abstract: Vector quantization techniques are currently favored in speech compression. Recently a nonadaptive form of vector quantization was proposed for monochrome image compression. In this paper, color picture compression by adaptive vector quantization is presented. Both spatial and spectral redundancy is exploited. Adaptive vector quantization of color pictures is shown to produce better mean-square-error result than block cosine transform coding, for the same data rate. No statistical image model is assumed. Decoding is a simple process of table look-up, permitting video-rate decoding, thus only a small refresh memory containing the codewords is necessary.

Abstract: In this paper we present results of a series of experiments in which combinations of vector quantization and temporal energy contours are incorporated into the standard framework for the word recognizer. We consider two distinct word vocabularies, namely, a set of 10 digits, and a 129-word airlines vocabulary. We show that the incorporation of energy leads to small but consistent improvements in performance for the digits vocabulary; the incorporation of vector quantization (in a judicious manner) leads to small degradation in performance for both vocabularies, but at the same time reduces overall computation of the recognizer by a significant amount. We conclude that a high-performance, moderate-computation, isolated word recognizer can be achieved using vector quantization and the temporal energy contour.

Abstract: The principles of classification applied to the representation of the words in a vocabulary lead to the clustering of the acoustic vectors into prototype vectors. For a small number of prototypes, recognition scores comparable to those observed with unclustered vocabularies are obtained with a highly reduced computation time. Two different forms (deterministic and stochastic) of the single-level recognition method for concatenated words are described and the improvements obtained by vector quantization are put into evidence. The use of prototypes in the training phase of the finite stochastic automata representing a vocabulary word is also described.

Abstract: The theory of vector quantization (VQ) of linear predictive coding (LPC) coefficients has established a wide variety of techniques for quantizing LPC spectral shape to minimize overall spectral distortion. Such vector quantizers have been widely used in the areas of speech coding and speech recognition. The conventional vector quantizer utilizes only spectral shape information and essentially disregards the energy or gain term associated with the optimal LPC fit to the signal being modeled. In this paper we present a method of incorporating LPC spectral shape and energy into the code-book entries of the vector quantizer. To do this, we postulate a distortion measure for comparing two LPC vectors that uses the weighted sum of an LPC shape distortion and a log energy distortion. Based on this combined distortion measure, we have designed and studied vector quantizers of several sizes for use in isolated word speech recognition experiments. We found that a fairly significant correlation exists between LPC shape and signal energy. Hence, an LPC shape combined with energy vector quantizer with a given distortion requires far fewer code-book entries than one in which LPC shape and energy are quantized separately. Based on isolated word recognition tests on both a 10-digit and a 129-word airlines vocabulary, we found improvements in recognition accuracy by using the VQ with both LPC shape and energy over that obtained using a VQ with LPC shape alone.

Abstract: A real-time full search vector quantization system for speech waveform coding is implemented using LSTTL and CMOS devices. The system consists of low-pass filters, A/D and D/A converters, an algorithm for discriminating voiced and unvoiced speed, a full search vector quantizer encoder and decoder, and a microprocessor-based controller. The system is designed to operate at two possible rates: one bit/sample using a dimension 8 vector quantizer (6500 bits/s) or 2 bits/sample using a dimension 4 vector quantizer (13 000 bits/s). In both cases the codebooks have rate 8 bits/vector. Separate codebooks were designed for voiced and unvoiced speech based on a training sequence of 640 000 samples containing five different speakers. The subjective and quantitative results are compared to both simulations and with a real-time array processor based implementation.

Abstract: The technique of vector quantization has been widely applied in the area of speech coding and has recently been introduced into the area of speech recognition. For the conventional statistical pattern recognition word recognizer using LPC feature sets as the analysis frames, the use of vector quantization leads to a large reduction in computation for the dynamic time warping pattern matching, and a concomittant small increase in average word error rate. A second technique that has been recommended for improving the performance of isolated word recognizers is the addition of temporal energy information into the distance metric for comparing frames of speech. It has been shown that the information in the prosodic energy contour complements the segmental information of the LPC spectrum, thereby providing small but consistent improvements in performance for small word vocabularies. In this talk we present results of a series of speaker independent, isolated word recognition tests using a 10‐word digits vocabu...

Abstract: A new interframe coding technique is proposed. Where a two-dimensional Hadamard transform is applied on each sub-block of successive frames, and an adaptive vector quantisation scheme is applied along the transformed blocks of the successive frames. The performance of the algorithm is evaluated by computer simulation on sequence of moving images.

Abstract: The theory of vector quantization (VQ) of linear predictive coding (LPC) coefficients has established a wide variety of techniques for quantizing LPC spectral shape to minimize overall spectral distortion. Such vector quantizers have been widely used in the areas of speech coding and speech recognition. The conventional vector quantizer utilizes only spectral shape information and essentially disregards the energy or gain term associated with the optimal LPC fit to the signal being modelled. In this paper we present a method of incorporating LPC spectral shape and energy into the codebook entries of the vector quantizer. To do this we postulate a distortion measure for comparing two LPC vectors which uses a weighted sum of an LPC shape distortion and a log energy distortion. Based on this combined distortion measure we have designed and studied vector quantizers of several sizes for use in isolated word speech recognition experiments. We have found that a fairly significant correlation exists between LPC shape and signal energy; hence a combined LPC shape plus energy vector quantizer with a given distortion requires far fewer codebook entries than one in which LPC shape and energy are quantized separately. Based on isolated word recognition tests on both a 10-digit and a 129 word airlines vocabulary, we have found improvements in recognition accuracy by using the VQ with both LPC shape and energy over that obtained using a VQ with LPC shape alone.

Abstract: An end-point detector for LPC speech using squared prediction error look-ahead and automatic/manual threshold determination is described. The detector is algorithmically simple, computationally efficient,and uses only one decision parameter. Preliminary tests indicate that it is relatively immune to transient pulses and various low-level noises, yet preserves low-level speech sounds such as weak fricatives to a significant extent under moderate noise conditions. Tests indicate that 93.8% of automatically determined endpoints agree to within two frames of manually determined endpoints. The detector is especially suitable for use in vector-quantization based LPC systems, where the squared prediction error is easily available.

Abstract: In many applications it is necessary to digitize data, knowing only that later on some random function of the digitized data will be of interest. This problem is investigated when the data digitizers are allowed to be multidimensional, i.e., they map a K -dimensional data vector into one of a set of K -dimensional output vectors. It is shown that very complex distortion measures arise naturally. Results are given for the error measure defined as the squared value of the difference between the function of the digitized data and the function of the undigitized data.

Abstract: Vector Quantization (VQ) and subband coding (SBC) are two of the most efficient data compression systems in the field of medium-to-low rate speech waveform coding. In this paper, we examine the performance of a coding system operating at either 6,500 bits per second or 13,000 bits per second which incorporates a full-search vector quantizer to encode the outputs coming from a complete subband coder filter bank.

Abstract: The DPCM (differential pulse code modulation) for color television signal transmission has been improved both in picture quality and information compression, by a vector signal quantization called embedding quantization. According to the present invention, vector signal comprising a plurality of elements is quantized while the quantized output signal is obtained after a plurality of quantization and prediction operations. Number of repetition times of the quantization and prediction is for instance equal to the number of elements composing vector signal. A vector is for instance composed of three elements relating to red, green, and blue primary signals of a color television signal.

Abstract: The use of vector quantization (VQ) in isolated-word recognition of a 20-word vocabulary is examined. A separate sequence of VQ code books is designed for each word in the recognition vocabulary and input words are classified by performing VQ and finding the sequence of code books that achieve the smallest average distortion. In this paper, critical parameters are noted and the results of parameter studies are presented.

Abstract: Two types of isolated digit recognition systems based on vector quantization were tested in a speaker-independent task. In both types of systems, a digit was modelled as a sequence of codebooks generated from segments of training data. In systems of the first type, the training and unknown utterances were simply partitioned into 1, 2 or 3 equal-length segments. Recognition involved computing the distortion when the input spectra were vector quantized using the codebook sequences. These systems are closely related to recognizers proposed by Burton et al.[1]. In systems of the second type, training segments corresponded to acoustic-phonetic units and were obtained from hand-marked data. Recognition involved generating a minimum-distortion segmentation of the unknown by dynamic programming. Accuracies approaching 96-97% were achieved by both types of systems.

Abstract: This paper describes a speech coder which exploits the combination of piecewise LPC analysis, full residual excitation and vector quantization in order to yield very good quality at 9.6 kbit/s. The piecewise approximation of the speech spectrum permits a higher spectral accuracy than standard LPC and reduces the computational load by about 40%. The full residual excitation overcomes the disadvantages of the base-band model, previously used at this bitrate. Finally, the vector quantization approach permits a dramatic bit saving over scalar quantization for LPC parameter compression, and provides a better distortion performance in the residual representation.