Large external memory bandwidth requirement leads to increased system power dissipation and cost in video coding application. Majority of the external memory traffic in video encoder is due to reference data accesses. We describe a lossy reference frame compression technique that can be used in video coding with minimal impact on quality while significantly reducing power and bandwidth requirement. The low cost transformless compression technique uses lossy reference for motion estimation to reduce memory traffic, and lossless reference for motion compensation (MC) to avoid drift. Thus, it is compatible with all existing video standards. We calculate the quantization error bound and show that by storing quantization error separately, bandwidth overhead due to MC can be reduced significantly. The technique meets key requirements specific to the video encode application. 24-39% reduction in peak bandwidth and 23-31% reduction in total average power consumption are observed for IBBP sequences.

The recently proposed H.264 video coding standard offers significant coding gains over previously defined standards. An enhanced intra-frame prediction algorithm has been proposed in H.264 for efficient compression of I-frames. This paper investigates the scope of the intraframe coder of H.264 for image coding. We compare the quality of this coder and the complexity of its decoder with the commonly used image codecs (JPEG and JPEG2000). Our results demonstrate that H.264 has a strong potential as an alternative to JPEG and JPEG2000.

A novel paradigm based on pitch-adaptive windows is proposed for solving the problem of encoding the fixed codebook excitation in low bit-rate CELP coders. In this method, the nonzero excitation in the fixed codebook is substantially localized to a set of time intervals called windows. The positions of the windows are adaptive to the pitch peaks in the linear prediction residual signal. Thus, high coding efficiency is achieved by allocating most of the available FCB bits to the perceptually important segments of the excitation signal. The pitch-adaptive method is adopted in the design of a novel multimode variable-rate speech coder applicable to CDMA-based cellular telephony. Results demonstrate that the adaptive windows method yields excellent voice quality and intelligibility at average bit-rates in the range of 2.5-4.0 kbps.

Many conventional speech recognition systems are based on the use of hidden Markov models (HMM) within the context of discriminant-based pattern classification. While the speech recognition objective is a low rate of misclassification, HMM design has been traditionally approached via maximum likelihood (ML) modeling which is, in general, mismatched with the minimum error objective and hence suboptimal. Direct minimization of the error rate is difficult because of the complex nature of the cost surface, and has only been addressed previously by discriminative design methods such as generalized probabilistic descent (GPD). While existing discriminative methods offer significant benefits, they commonly rely on local optimization via gradient descent whose performance suffers from the prevalence of shallow local minima. As an alternative, we propose the deterministic annealing (DA) design method that directly minimizes the error rate while avoiding many poor local minima of the cost. The DA is derived from fundamental principles of statistical physics and information theory. In DA, the HMM classifier's decision is randomized and its expected error rate is minimized subject to a constraint on the level of randomness which is measured by the Shannon entropy. The entropy constraint is gradually relaxed, leading in the limit of zero entropy to the design of regular nonrandom HMM classifiers. An efficient forward-backward algorithm is proposed for the DA method. Experiments on synthetic data and on a simplified recognizer for isolated English letters demonstrate that the DA design method can improve recognition error rates over both ML and GPD methods.

A new learning algorithm is proposed for piecewise regression modeling. It employs the technique of deterministic annealing to design space partition regression functions. While the performance of traditional space partition regression functions such as CART and MARS is limited by a simple tree-structured partition and by a hierarchical approach for design, the deterministic annealing algorithm enables the joint optimization of a more powerful piecewise structure based on a Voronoi partition. The new method is demonstrated to achieve consistent performance improvements over regular CART as well as over its extension to allow arbitrary hyperplane boundaries. Comparison tests, on several benchmark data sets from the regression literature, are provided.

We attack the general problem of HMM-based speech recognizer design, and in particular, the problem of isolated letter recognition in the presence of background noise. The standard design method based on maximum likelihood (ML) is known to perform poorly when applied to isolated letter recognition. The minimum classification error (MCE) approach directly targets the ultimate design criterion and offers substantial improvements over the ML method. However, the standard MCE method relies on gradient descent optimization which is susceptible to shallow local minima traps. We propose to overcome this difficulty with a powerful optimization method based on deterministic annealing (DA). The DA method minimizes a randomized MCE cost subject to a constraint on the level of entropy which is gradually relaxed. It may be derived based on information-theoretic or statistical physics principles. DA has a low implementation complexity and outperforms both standard ML and the gradient descent based MCE algorithm by a factor of 1.5 to 2.0 on the benchmark CSLU spoken letter database. Further, the gains are maintained under a variety of background noise conditions.

A new and effective design method is presented for statistical regression functions that belong to the class of mixture models. The class includes the hierarchical mixture of experts (HME) and the normalized radial basis functions (NRBF). Design algorithms based on the maximum likelihood (ML) approach, which emphasize a probabilistic description of the model, have attracted much interest in HME and NRBF models. However, their design objective is mismatched to the original squared-error regression cost and the algorithms are easily trapped by poor local minima on the cost surface. In this paper, we propose an extension of the deterministic annealing (DA) method for the design of mixture-based regression models. We construct a probabilistic framework, but unlike the ML method, we directly optimize the squared-error regression cost, while avoiding poor local minima. Experimental results show that the DA method outperforms standard design methods for both HME and NRBF regression models.

In vector quantization, one approximates an input random vector, Y, by choosing from a finite set of values known as the codebook. We consider a more general problem where one may not have direct access to Y but only to some statistically related random vector X. We observe X and would like to generate an approximation to Y from a codebook of candidate vectors. This operation, called generalized vector quantization (GVQ), is essentially that of quantized estimation. An important special case of GVQ is the problem of noisy source coding wherein a quantized approximation of a vector, Y, is obtained from observation of its noise-corrupted version, X. The optimal GVQ encoder has high complexity. We overcome the complexity barrier by optimizing a structurally-constrained encoder. This challenging optimization task is solved via a probabilistic approach, based on deterministic annealing, which overcomes problems of shallow local minima that trap simpler descent methods. We demonstrate the successful application of our method to the coding of noisy sources.

In many signal compression applications, the evolution of the signal over time can be represented by a sequence of random vectors with varying dimensionality. Frequently, the generation of such variable-dimension vectors can be modeled as a random sampling of another signal vector with a large but fixed dimension. Efficient quantization of these variable-dimension vectors is a challenging task and a critical issue in speech coding algorithms based on harmonic spectral modeling. We introduce a simple and effective formulation of the problem and present a novel technique, called variable-dimension vector quantization (VDVQ), where the input variable-dimension vector is directly quantized with a single universal codebook. The application of VDVQ to low bit-rate speech coding demonstrates significant gain in subjective quality as well as in rate-distortion performance over prior indirect methods.

Optimal vector quantization of variable-dimension vectors in principle is feasible by using a set of fixed dimension VQ codebooks. However, for typical applications, such a multi-codebook approach demands a grossly excessive and impractical storage and computational complexity. Efficient quantization of such variable-dimension spectral shape vectors is the most challenging and difficult encoding task required in an important family of low bit-rate vocoders. The authors introduce a simple and effective formulation of variable-dimension vector quantization (VDVQ) which quantizes variable-dimension vectors using a single universal codebook having fixed dimension yet covering the entire range of input vector dimensions under consideration. This VDVQ technique is applied to quantize variable-dimension spectral shape vectors leading to a high quality speech coder at the low bit-rate of 2.5 kb/s. The combination of a universal spectral codebook and structured VQ reduces storage and computational complexity, yet delivers a high quantization efficiency and enhanced perceptual quality of the coded speech.>