Kevin T. Malone

Enumeration and trellis-searched coding schemes for speech LSP parameters

Speech coders employing forward adaptive predictive coding (APC) and operating at medium-to-low bit rates necessitate efficient encoding of the linear predictive coding (LPC) coefficients. Line spectrum pair (LSP) parameters are currently one of the most efficient choices of transmission parameters for the LPC coefficients. The authors briefly reviews LSP parameters and presents several low-delay coding schemes for the parameters. The coders are simulated using data generated from both the autocorrelation and covariance LPC analysis methods. The performances of the coders are given for a variety of rates and LPC analysis conditions. The most efficient scheme developed uses a uses a predictive form of trellis-coded quantization (TCQ). Its performance is comparable or superior to that of other low-delay LSP coding schemes. An enumeration scheme that reduces the rate of a given scalar quantization structure without decreasing coder performance is also presented. >

A Sequential Vehicle Classifier for Infrared Video using Multinomial Pattern Matching

Vehicle classification is a challenging problem, since vehicles can take on many different appearances and sizes due to their form and function, and the viewing conditions. The low resolution of uncooled-infrared video and the large variability of naturally occurring environmental conditions can make this an even more difficult problem. We develop a multilook fusion approach for improving the performance of a single look system. Our single look approach is based on extracting a signature consisting of a histogram of gradient orientations from a set of regions covering the moving object. We use the multinomial pattern matching algorithm to match the signature to a database of learned signatures. To combine the match scores of multiple signatures from a single tracked object, we use the sequential probability ratio test. Using real infrared data we show excellent classification performance, with low expected error rates, when using at least 25 looks.

Trellis-searched adaptive predictive coding of speech

Trellis coded vector quantization (TCVQ) and forward adaptive predictive coding (APC) are used to form an efficient speech coding system operating at bit rates of 16 and 9.6 kb/s. The effectiveness of the system is studied for a variety of system parameters and utterances. Simulation results indicate that segmental signal-to-noise ratios as high as 23.8 and 15.4 dB are obtainable at 16 and 9.6 kb/s respectively. The quality of the reconstructed speech is deemed to be excellent at 16 kb/s and very good at 9.6 kb/s. An algorithm for optimizing the residual codebooks is presented. >

Contour-gain vector quantization

An information-theory motivation for considering a contour-gain codebook structure is given and an iterative contour-gain vector quantizer (CGVQ) algorithm allowing optimization of the shape codebook for a fixed-gain codebook is described. Numerical results are presented for CGVQ encoding of first-, second-, and tenth-order Gauss-Markov sources, and a clear improvement over SGVQ (shape-gain vector quantizer) performance is demonstrated. Numerical results are also presented for CGVQ waveform encoding of speech, and again an improvement over SGVQ performance is demonstrated. The perceptual quality of the encoded speech was roughly equivalent for the two models. >

Transform Coding of Speech with Pyramid Vector Quantization

The pyramid vector quantizer (PVQ) is a lattice quantizer that was motivated bv the geometric properties of a memoryless Laplacian source. For large rates and the cubic lattice. the PVQ provides improvements of 2.39, 5.64, and 8.40 dB for memoryless Gaussian, Laplacian, and gamma sources, respectivelv, compared to the corresponding optimum (noniuniform) scalar quantizer. The lattice basis of the PVQ allows simple encoding and decoding algorithms with a complexity that grows only linearly with the vector dimension. A correlated source such as speech has a geometric nature that is not well suited to the PVQ unless transform coding is used. It is demonstrated that an encoding system using a cosine transform, interleaving of the transform coefficients, and pyramid vector quantization can achieve signal-to-noise ratio (SNR) performance in excess of 20dB for an average rate of 2 bits/sample.

Classifying Acoustic Signatures Using the Sequential Probability Ratio Test

Abstract Acoustic sensors can provide real time information about moving targets. The acoustic information is typically processed sequentially, allowing the sequential probability ratio test (SPRT) to be used as the basis to solve the target identification problem. The SPRT keeps gathering observations only as long as the statistical test has a value between the upper stopping boundary and the lower stopping boundary. When the test goes above the upper boundary or below the lower boundary, the system can make a decision. The desired false alarm error rate and the desired missed detection error rate determine the upper and lower stopping boundaries. We present extensions to the sequential probability ratio test to handle problems of dependence, contamination, and the unknown class. We also present results for using the SPRT for target identification using acoustic information.

Trellis-searched adaptive predictive coding

Considers the use of TCO (trellis coded quantization) in an adaptive predictive coding (APC) structure at encoding rates slightly above one and two bits per sample. The authors examine the effect of long- and short-term predictor order on the encoding performance and study the variation in performance as a function of the TCQ codebook. The simulation results indicate that SNRs (signal/noise ratio) and SEGSNRs (segmental SNRs) as high as 25 and 24 dB, respectively, are possible at rates slightly above two bits per sample using a simple four-state trellis. At higher near one bit per sample, SNRs and SEGNRs as high as 16.5 and 15.5 dB, respectively, are possible.<<ETX>>

Vehicle Classification in Infrared Video Using the Sequential Probability Ratio Test

This chapter develops a multilook fusion approach for improving the performance of a single-look vehicle classification system for infrared video. Vehicle classification is a challenging problem since vehicles can take on many different appearances and sizes due to their form and function and the viewing conditions. The low resolution of uncooled infrared video and the large variability of naturally occurring environmental conditions can make this an even more difficult problem. Our single-look approach is based on extracting a signature consisting of a histogram of gradient orientations from a set of regions covering the moving object. We use the multinomial pattern-matching algorithm to match the signature to a database of learned signatures. To combine the match scores of multiple signatures from a single tracked object, we use the sequential probability ratio test. Using infrared data, we show excellent classification performance, with low expected error rates, when using at least 25 looks.

Acoustic/seismic identifications, imaging, and communications in Steel Rattler

Unattended ground sensors (UGS) utilize data from a variety of sensors (e.g., acoustic, seismic, and imagery) to make a determination about an unknown potential target. The Steel Rattler UGS derives its target identification solution from acoustic and seismic data. The identification solution and optional still imagery of the target are transmitted to the appropriate operating bases via satellite. This paper describes the various Steel Rattler hardware components used in the target identification process, the optional imaging system, and the communication system used for testing and demonstration purposes.

Markov sequential pattern recognition : dependency and the unknown class.

The sequential probability ratio test (SPRT) minimizes the expected number of observations to a decision and can solve problems in sequential pattern recognition. Some problems have dependencies between the observations, and Markov chains can model dependencies where the state occupancy probability is geometric. For a non-geometric process we show how to use the effective amount of independent information to modify the decision process, so that we can account for the remaining dependencies. Along with dependencies between observations, a successful system needs to handle the unknown class in unconstrained environments. For example, in an acoustic pattern recognition problem any sound source not belonging to the target set is in the unknown class. We show how to incorporate goodness of fit (GOF) classifiers into the Markov SPRT, and determine the worse case nontarget model. We also develop a multiclass Markov SPRT using the GOF concept.