Mutual Information, the Linear Prediction Model, and CELP Voice Codecs

We write the mutual information between an input speech utterance and its reconstruction by a code-excited linear prediction (CELP) codec in terms of the mutual information between the input speech and the contributions due to the short-term predictor, the adaptive codebook, and the fixed codebook. We then show that a recently introduced quantity, the log ratio of entropy powers, can be used to estimate these mutual informations in terms of bits/sample. A key result is that for many common distributions and for Gaussian autoregressive processes, the entropy powers in the ratio can be replaced by the corresponding minimum mean squared errors. We provide examples of estimating CELP codec performance using the new results and compare these to the performance of the adaptive multirate (AMR) codec and other CELP codecs. Similar to rate distortion theory, this method only needs the input source model and the appropriate distortion measure.

Mutual Information, the Linear Prediction Model, and CELP Voice Codecs

We write the mutual information between an input speech utterance and its reconstruction by a Code-Excited Linear Prediction (CELP) codec in terms of the mutual information between the input speech and the contributions due to the short term predictor, the adaptive codebook, and the fixed codebook. We then show that a recently introduced quantity, the log ratio of entropy powers, can be used to estimate these mutual informations in terms of bits/sample. A key result is that for many common distributions and for Gaussian autoregressive processes, the entropy powers in the ratio can be replaced by the corresponding minimum mean squared errors. We provide examples of estimating CELP codec performance using the new results and compare to the performance of the AMR codec and other CELP codecs. Similar to rate distortion theory, this method only needs the input source model and the appropriate distortion measure.

Secret Communication Techniques

This chapter introduces the reader to cryptography, steganography, watermarking, and quality parameters of image steganography techniques. Cryptography is a technique for secret communication. Steganography is a technique for secret and unnoticeable communication. The watermarking techniques hide watermarks inside the digital media. There are four types of steganography techniques: (1) image steganography, (2) audio steganography, (3) video steganography, and (4) text steganography. The quality of image steganographic algorithms can be measured by three parameters like (1) hiding capacity, (2) distortion measure, and (3) security check.

High Capacity Image Steganography Using Modified LSB Substitution and PVD against Pixel Difference Histogram Analysis

The combination of pixel value differencing (PVD) and least significant bit (LSB) substitution gives higher capacity and lesser distortion. However, there are three issues to be taken into account: (i) fall off boundary problem (FOBP), (ii) pixel difference histogram (PDH) analysis, and (iii) RS analysis. This paper proposes a steganography technique in two variants using combination of modified LSB substitution and PVD by taking care of these three issues. The first variant operates on 2 × 3 pixel blocks and the second technique operates on 3 × 3 pixel blocks. In one of the pixels of a block, embedding is performed using modified LSB substitution. Based on the new value of this pixel, difference values with other neighboring pixels are calculated. Using these differences, PVD approach is applied. The edges in multiple directions are exploited, so PDH analysis cannot detect this steganography. The LSB substitution is performed in only one pixel of the block, so RS analysis also cannot detect this steganography. To address the FOBP, suitable equations are used during embedding procedure. The experimental results such as bit rate and distortion measure are satisfactory.

Regularized sparse decomposition model for speech enhancement via convex distortion measure

An important stage in speech enhancement is to estimate noise signal which is a difficult task in non-stationary and low signal-to-noise conditions. This paper presents an iterative speech enhancement approach which requires no prior knowledge of noise and is based on low-rank sparse matrix decomposition using Gammatone filterbank and convex distortion measure. To estimate noise and speech, the noisy speech is decomposed into low-rank noise and sparse-speech parts by enforcing sparsity regularization. The exact distribution of noise signals and noise estimator is not required in this approach. The experimental results demonstrate that our approach outperforms competing methods and yields better overall speech quality and intelligibility. Moreover, composite objective measure reinforced a better performance in terms of residual noise and speech distortion in adverse noisy conditions. The time-varying spectral analysis validates significant reduction of the background noise.