Speech Technologies for Language Therapy

This chapter presents the development and use of speech technologies in language therapy for children with hearing disabilities. It describes the challenges that must be addressed to design and construct a system to support effective interactions. The chapter begins with an introduction to speech and language therapy and discusses how speech-based systems can provide useful tools for speech and language therapy and to overcome the lack of sufficient human resources to help all children who require it. Then it describes the construction of adequate speech recognition systems for children, using artificial neural networks and hidden Markov models. Next, a case study is presented with the system we have been developing for speech and language therapy for children in a special education school. The chapter concludes with an analysis of the obtained results and the lessons learned from our experiences that will hopefully inform and encourage other researchers, developers, and educators to develop learning tools for individuals with disabilities.

Differentially Fed Artificial Neural Networks for Speech Signal Prediction

Speaker authentication has become increasingly important. It goes with the other forms of security checks such as user login and personal identification number and has a say in the final decision about the authenticity. One of the issues with the authentication algorithms is that the automated devices that take the call have to work with a limited data set. In this chapter, a new class of intelligent element called differentially fed artificial neural network has been introduced to predict the data and use it effectively. It keeps the model simple and helps in taking online and crisp decisions with the available limited data.

Spectral-Based Analysis and Synthesis of Audio Signals

This chapter reviews audio signal processing techniques related to sound generation via additive synthesis. Particular focus will be put on sinusoidal modeling. Each processing stage involved in obtaining a sinusoidal representation for audio signals is described. Then, synthesis techniques that allow reconstructing an audio signal based on a given parametric representation are presented. Finally, some audio applications where sinusoidal modeling is employed are briefly discussed.

Introduction to Speech Recognition

This chapter presents the state-of-the-art automatic speech recognition (ASR) technology, which is a very successful technology in the computer science field, related to multiple disciplines such as the signal processing and analysis, mathematical statistics, applied artificial intelligence and linguistics, and so forth. The unit of essential information used to characterize the speech signal in the most widely used ASR systems is the phoneme. However, recently several researchers have questioned this representation and demonstrated the limitations of the phonemes, suggesting that ASR with better performance can be developed replacing the phoneme by triphones and syllables as the unit of essential information used to characterize the speech signal. This chapter presents an overview of the most successful techniques used in ASR systems together with some recently proposed ASR systems that intend to improve the characteristics of conventional ASR systems.

Advanced Techniques in Speech Recognition

During the last 30 years, people have tried to communicate in an oral form with the computers, developing for this end an important amount of automatic speech recognition algorithms. Because of this, software such as the Dragon Dictate and the IBM Via Voice are already available to interact easily with the computer in oral form. However, during the last several years ASR has not reported important advances, not only due to the advances obtained until now, but also because the scientific community working in this area does not have founded another tool so powerful as HMM, despite a great number of alternatives that have been proposed since HMM appeared. This chapter presents the main elements required to create a practical ASR using HMM. The basic principles of the continuous density hidden Markov models (CDHMM) are also given.

Adaptive Filters

This chapter focuses on the main aspects of adaptive signal processing. The basic concepts are introduced in a simple framework, and its main applications (namely system identification, channel equalization, signal prediction, and noise cancellation) are briefly presented. Several adaptive algorithms are presented, and their convergence behaviors are analyzed. The algorithms considered in this chapter include the popular least-mean square (LMS), its normalized-LMS version, the affine-projection with the set-membership variation, the recursive least-squares (RLS), the transform-domain, the sub-band domain, and some IIR-filter algorithms such as the equation-error (EE) and the output-error (OE) algorithms. The main purpose of all this presentation is to give general guidelines for the reader to choose the most adequate technique for the audio application at hand.

DSP Techniques for Sound Enhancement of Old Recordings

This chapter addresses digital signal processing techniques for sound restoration and enhancement. The most common sound degradations found in audio recordings, such as thumps, pops, clicks, and hiss are characterized. Moreover, the most popular solutions for sound restoration are described, with emphasis on their practical applicability. Finally, critical views on the performance of currently available restoration algorithms are provided, along with discussions on new tendencies observed in the field.

Active Noise Canceling

Some of the main problems present in active noise cancellations are the feedback distortion due to the acoustic feedback between the cancellation speaker and the input microphone; the high computational complexity when recursive least square algorithms are used; and the secondary path estimation. This chapter presents a review of some successful solutions to these problems, such as a hybrid structure to reduce the feedback distortion; active noise cancellation (ANC) FxRLS algorithms in which the filter input signal is decomposed into a finite number; and mutually near orthogonal signal components, as well as successful secondary path estimation algorithms. Computer simulation results confirm the desirable properties of presented ANC structures.

Adaptive Digital Filtering and Its Algorithms for Acoustic Echo Canceling

In this chapter, we shall study adaptive digital filtering (ADF) and its application to acoustic echo canceling (AEC). At first, Wiener filtering and algorithms such as LMS in the time domain for ADF are explained. Then, to decrease the computational complexity, the frequency domain algorithms such as FDAF and FBAF will be studied. To challenge the double-talk problem in AEC, we will also introduce various algorithms by processing the correlation function of the signal. The proposed algorithms here are CLMS, ECLMS, and using frequency domain is FECLMS, and using wavelet transform is WECLMS. Each of these algorithms has its own merits, and they will be evaluated. At the end of this chapter a new system for room-acoustic partitioning is proposed. This new system is called smart acoustic room (SAR). The SAR will also be used in AEC with double-talk condition. The authors wish to gather all aspects in studying ADF and their use in AEC by going very deep into theoretical details as well as considering more practical and feasible applications considering real-time implementation.

Digital Watermarking Techniques for Audio and Speech Signals

The main objective of this chapter is to provide an overview of existing speech and audio watermarking technology and to demonstrate the importance of signal processing for the design and evaluation of watermarking algorithms. This chapter describes the factors to be considered while designing speech and audio watermarking algorithms, including the choice of the domain and signal features for watermarking, watermarked signal fidelity, watermark robustness, data payload, security, and watermarking applications. The chapter presents several state-of-the-art speech and audio watermarking algorithms and discusses their advantages and disadvantages. The various applications of watermarking and developments in performance evaluation of watermarking algorithms are also described.