Predictive Coding System Research Articles

Survey of Adaptive Image Coding Techniques

The following is a survey of the technical literature on adaptive coding of imagery. Section 1 briefly discusses the general problem of image data compression. The optimum image data compression system, from a theoretical viewpoint, is presented in Section 1.1. The theoretical and the implementational problems of the optimum system are discussed and the assumptions that are made to overcome these problems are outlined. One important assumption is the stationarity which is not true for most imagery. In adaptive systems the parameters are varied according to changes in signal statistics optimizing the system performance for nonstationary signals. In this article the adaptive systems are divided to four categories. Section 2 is a survey of adaptive transform coding systems. Section 3 discusses adaptive predictive coding systems. Sections 4 and 5 discuss adaptive cluster coding and adaptive entropy technique, respectively.

Voice-excited predictive coding system for low-bit-rate transmission of speech

Voice-excited speech analysis synthesis systems eliminate two of the most difficult tasks in speech analysis, namely, voiced-unvoiced decision and pitch analysis. They can therefore perform better than pitch-excited systems under practical speaking conditions. However, the transmission rates for voice-excited systems are usually much higher than for pitch-excited systems, owing to the additional bits needed for encoding the baseband. This disadvantage can be overcome by efficient encoding of the baseband. This paper presents a voice-excited predictive coding method for speech transmission at 2400 to 3600 bps. The analytic signal corresponding to a baseband of 300 to 900 Hz is square-rooted to reduce its bandwidth to 300 Hz and is encoded at 1200 bps by predictive coding. At the receiver, the wide-band excitation is derived by nonlinear distortion and inverse-filtering of the baseband. The speech signal is sampled at 6.67 kHz, and its spectral-envelope information is transmitted every 20 msec by 10 area parameters obtained by the linear prediction method. These parameters and the rms value of the speech signal are encoded at 1800 bps, yielding a transmission rate of 3000 bps for the speech signal.

Real-Time Autocorrelation Predictive Coder

We describe a real-time autocorrelation predictive coding system for speech compression that has been implemented on a programmable signal processor having a word size of 16 bits and a multiply time of 750 nsec. This system employs the normal equations to generate an all-pole filter which, when impulsed, produces the least-squares approximation to the input signal. The technique uses a 6.4-kHz sampling rate on an input bandwidth of 3 kHz, a lOth-order predictive filter, an average magnitude difference function (AMDF) pitch extractor, and pitch synchronous interpolation of the gain and pitch parameters to produce transmission rates of 2.4 and 3.6 k bits/sec. The paper emphasizes the modifications needed to achieve real-time operation and the accuracy and timing requirements to extract pitch, calculate the autocorrelation and filter coefficients, and synthesize the output.

Implementation of a Linear Predictive Coder

This paper will describe the design and implementation approach of a predictive coding system for speech compression based, in part, on the autocorrelation method [F. Itakura, IEEE Trans. Comm. COM-20, 4 (1972)], a computationally more efficient technique than the autocovariant method [B. S. Atal, J. Acoust. Soc. Amer. 50, 2 (1971), but which produces speech of similar quality. The design requires a 6400-Hz sampling rate on an input bandwidth of 3 kHz, a 10th-order predictive filter, an absolute magnitude difference function (AMDF) pitch extractor, and a pitch-synchronous interpolation of the synthesizer parameters to produce transmission rates of 2.4 and 4.8 k bps. The computational procedures to implement the algorithms with fixed-point arithmetic on a processor having a word size of 16 bits will be discussed with special reference given toward (a) calculating the first 11 autocorrelation functions, (b) inverting the 10×10 Toeplitz matrix and solving for the 10 filter coefficients, (c) computing gain and pitch and interpolating synthesizer parameter pitch synchronously, and (d) synthesizing speech with a recursive lattice filter. System performance will be demonstrated.

Adaptive Predictive Coding of Speech Signals

We describe in this paper a method for efficient encoding of speech signals, based on predictive coding. In this coding method, both the transmitter and the receiver estimate the signal's current value by linear prediction on the previously transmitted signal. The difference between this estimate and the true value of the signal is quantized, coded and transmitted to the receiver. At the receiver, the decoded difference signal is added to the predicted signal to reproduce the input speech signal. Because of the nonstationary nature of the speech signals, an adaptive linear predictor is used, which is readjusted periodically to minimize the mean-square error between the predicted and the true value of the signals. The predictive coding system was simulated on a digital computer. The predictor parameters, comprising one delay and nine other coefficients related to the signal spectrum, were readjusted every 5 milliseconds. The speech signal was sampled at a rate of 6.67 kHz, and the difference signal was quantized by a two-level quantizer with variable step size. Subjective comparisons with speech from a logarithmic PCM encoder (log-PCM) indicate that the quality of the synthesized speech signal from the predictive coding system is approximately equal to that of log-PCM speech encoded at 6 bits/sample. Preliminary studies suggest that the binary difference signal and the predictor parameters together can be transmitted at approximately 10 kilobits/second which is several times less than the bit rate required for log-PCM encoding with comparable speech quality.