Browsing by Author "Rekik, Siwar"
Now showing 1 - 5 of 5
Results Per Page
Sort Options
Item An autoregressive time delay neural network for speech steganalysis(Institute of Electrical and Electronics Engineers Inc., 2012) Rekik, Siwar; Selouani, Sid-Ahmed; Guerchi, Driss; Hamam, HabibHiding a secret message in speech signal, called steganography, is used to provide secure communication. The detection of hidden information in the transmitted message called steganalysis. The purpose of steganalysis is to identify the presence of embedded information, and does not actually attempt to extract or decode the hidden data. An automated method is required for detecting the existence of hidden message, since the huge amount of channeled information. However, the development and evaluation of steganalysis algorithms is a challenging task. In this paper we advocate a new steganalysis technique to classify a speech as having hidden information or not, using a powerful and sophisticated classifier called Autoregressive Time Delay Neural Network (AR-TDNN). The originality of this AR-TDNN is its quite ability to detect secret messages hidden with different steganographic algorithms, although the variation of detection rate depends on the particular hiding techniques and amount of hidden information. © 2012 IEEE.Item A cooperative and conversational virtual agent for M-commerce applications(Institute of Electrical and Electronics Engineers Inc., 2009) Rekik, Siwar; Selouani, Sid-Ahmed; Hamam, HabibThe aim of this paper is to present a cooperative approach to improve the Human-System spoken dialogues. The main advantage of the proposed approach is its ability to reach both of the user and the system goals more efficiently. The strategy that underlines our system is well adapted to the mobile applications since it involves effective spoken exchanges to reach the users' and system mutual goals. The proposed fram ework is built in order to use the Hidden Markov Models (HMMs) based CMU-Sphinx speech recognition engine f or mobile communications. To evaluate our approach a case-study in the M-trading field is considered. The analysis of the case study shows the efficiency ofour strategy in comparison with the usual ones. ©2009 IEEE.Item Encoding spectral parameters using cache codebook(Springer Nature Switzerland AG, 2012) Guerchi, Driss; Rekik, SiwarA new efficient approach to quantize the spectral line frequencies (LSF) in a coder is proposed. The use of the full search algorithm in the spectral parameters quantization causes high complexity and large hardware storage. Attempts to reduce the complexity have been performed by lowering the size of the LSF codebook. This option leads to a sub-optimal solution; the number of LSF vectors to be tested affects the performance of the speech coder. Cache codebook (CCB) technique enhances the search of the optimal quantized spectral information. In this technique the size of the main codebook is kept unchanged while the number of closest match searches is reduced. Unlike the classical quantizer design, the CCB method involves one main codebook embedding four disjoint sub-codebooks. The content of the CCB at any time is an exact reproduction of one of the four sub-codebooks. The search for the best match to an input vector is limited to the LSF vectors of the CCB. Some criteria are used to accept or reject this closest match. The CCB is updated whenever the decision is in favor of rejection. The cache codebook was successfully embedded in a CELP coder to enhance the quantization of the spectral information. The comparison simulation results show that the Codebook Caching approach yields to comparable objective and subjective performance to that of the optimal full-search technique when using the same training and testing database. © Springer-Verlag Berlin Heidelberg 2012.Item On performing speech steganalysis using speech recognition and perceptual criteria(2010) Rekik, Siwar; Guerchi, Driss; Selouani, Sid-Ahmed; Hamam, HabibItem Speech steganography using wavelet and Fourier transforms(Springer Nature Switzerland AG, 2012) Rekik, Siwar; Guerchi, Driss; Selouani, Sid-Ahmed; Hamam, HabibA new method to secure speech communication using the discrete wavelet transforms (DWT) and the fast Fourier transform is presented in this article. In the first phase of the hiding technique, we separate the speech high-frequency components from the low-frequency components using the DWT. In a second phase, we exploit the low-pass spectral proprieties of the speech spectrum to hide another secret speech signal in the low-amplitude high-frequency regions of the cover speech signal. The proposed method allows hiding a large amount of secret information while rendering the steganalysis more complex. Experimental results prove the efficiency of the proposed hiding technique since the stego signals are perceptually indistinguishable from the equivalent cover signal, while being able to recover the secret speech message with slight degradation in the quality. © 2012 Rekik et al.; licensee Springer.