Sinha Speech Processing in Embedded Systems

1. Auflage 2009
ISBN: 978-0-387-75581-6
Verlag: Springer-Verlag
Format: PDF
Kopierschutz: Adobe DRM (»Systemvoraussetzungen)

E-Book, Englisch, 171 Seiten

Speech Processing has rapidly emerged as one of the most widespread and well-understood application areas in the broader discipline of Digital Signal Processing. Besides the telecommunications applications that have hitherto been the largest users of speech processing algorithms, several non-traditional embedded processor applications are enhancing their functionality and user interfaces by utilizing various aspects of speech processing.

"Speech Processing in Embedded Systems" describes several areas of speech processing, and the various algorithms and industry standards that address each of these areas. The topics covered include different types of Speech Compression, Echo Cancellation, Noise Suppression, Speech Recognition and Speech Synthesis. In addition this book explores various issues and considerations related to efficient implementation of these algorithms on real-time embedded systems, including the role played by processor CPU and peripheral functionality.

Autoren/Hrsg.

Sinha, Priyabrata

Inhaltsverzeichnis

1;Preface;5
2;Contents;7
3;Introduction;12
3.1;Digital vs. Analog Systems;12
3.2;Embedded Systems Overview;14
3.3;Speech Processing in Everyday Life;15
3.4;Common Speech Processing Tasks;16
3.5;Summary;18
3.6;References;18
4;Signal Processing Fundamentals;19
4.1;Signals and Systems [1];19
4.2;Sampling and Quantization [1, 2];21
4.3;Convolution and Correlation [2, 3];25
4.4;Frequency Transformations and FFT [1];30
4.5;Introduction to Filters [1];35
4.6;FIR and IIR Filters [1, 2];40
4.7;Interpolation and Decimation [1, 2];45
4.8;Summary;46
4.9;References;46
5;Basic Speech Processing Concepts;47
5.1;Mechanism of Human Speech Production [1, 3];47
5.2;Types of Speech Signals [1, 2];49
5.3;Digital Models for the Speech Production System [1, 2];52
5.4;Alternative Filtering Methodologies Used in Speech Processing [ 2, 4];53
5.5;Some Basic Speech Processing Operations [1, 2];57
5.6;Key Characteristics of the Human Auditory System [3];59
5.7;Evaluation of Speech Quality [2];61
5.8;Summary;63
5.9;References;64
6;CPU Architectures for Speech Processing;65
6.1;The Microprocessor Concept;65
6.2;Microcontroller Units Architecture Overview [3–5, 9, 10];67
6.3;Digital Signal Processor Architecture Overview [3–5, 7, 8];69
6.4;Digital Signal Controller Architecture Overview [3–5];70
6.5;Fixed-Point and Floating-Point Processors [1, 2, 6];70
6.6;Accumulators and MAC Operations [2, 3];72
6.7;Multiplication, Division, and 32-Bit Operations [3, 4];75
6.8;Program Flow Control [3, 4];76
6.9;Special Addressing Modes [3, 4];77
6.10;Data Scaling, Normalization, and Bit Manipulation Support [ 2– 4];80
6.11;Other Architectural Considerations [5, 6];81
6.12;Summary;84
6.13;References;84
7;Peripherals for Speech Processing;85
7.1;Speech Sampling Using Analog-to-Digital Converters [2, 3, 8–10];85
7.2;Speech Playback Using Digital-to-Analog Converters [2, 3];90
7.3;Speech Playback Using PulseWidth Modulation [3, 4];91
7.4;Interfacing with Audio Codec Devices [3, 8–10];92
7.5;Communication Peripherals [3, 5–7];95
7.6;Other Peripheral Features [1, 8–10];100
7.7;Summary;100
7.8;References;101
8;Speech Compression Overview;102
8.1;Speech Compression and Embedded Applications [2, 3, 5];102
8.2;Types of Speech Compression Techniques [1–4];105
8.3;Comparison of Speech Coders [5];106
8.4;Summary;108
8.5;References;109
9;Waveform Coders;110
9.1;Introduction to Scalar Quantization;110
9.2;Uniform Quantization [1–4];111
9.3;Logarithmic Quantization [1–4];112
9.4;ITU-T G.711 Speech Coder [5];113
9.5;ITU-T G.726 and G.726A Speech Coders [7, 8];114
9.6;ITU-T G.722 Speech Coder [6];117
9.7;Summary;119
9.8;References;121
10;Voice Coders;122
10.1;Linear Predictive Coding [1–4];122
10.2;Vector Quantization [1–4];127
10.3;Speex Speech Coder [8];128
10.4;ITU-T G.728 Speech Coder [5];129
10.5;ITU-T G.729 Speech Coder [6];131
10.6;ITU-T G.723.1 Speech Coder [7];131
10.7;Summary;133
10.8;References;133
11;Noise and Echo Cancellation;135
11.1;Benefits and Applications of Noise Suppression [4–8];135
11.2;Noise Cancellation Algorithms for 2-Microphone Systems;138
11.3;Noise Suppression Algorithms for 1-Microphone Systems [4];141
11.4;Active Noise Cancellation Systems [1, 7];143
11.5;Benefits and Applications of Echo Cancellation [1, 5, 6];144
11.6;Acoustic Echo Cancellation Algorithms [2, 5];146
11.7;Line Echo Cancellation Algorithms [3, 6];148
11.8;Computational Resource Requirements;148
11.9;Summary;149
11.10;References;150
12;Speech Recognition;151
12.1;Benefits and Applications of Speech Recognition [2, 3];151
12.2;Speech Recognition Using Template Matching [1, 3];155
12.3;Speech Recognition Using Hidden Markov Models [1, 3];158
12.4;Viterbi Algorithm [1, 3];159
12.5;Front-End Analysis [1, 3];160
12.6;Other Practical Considerations [1, 3];161
12.7;Performance Assessment of Speech Recognizers [2, 3];162
12.8;Computational Resource Requirements [3];162
12.9;Summary;163
12.10;References;163
13;Speech Synthesis;164
13.1;Benefits and Applications of Concatenative Speech Synthesis [2];164
13.2;Benefits and Applications of Text-to-Speech Systems [2];166
13.3;Speech Synthesis by Concatenation ofWords and Subwords [1];167
13.4;Speech Synthesis by Concatenating Waveform Segments [1];168
13.5;Speech Synthesis by Conversion from Text (TTS) [1];169
13.6;Computational Resource Requirements [1];171
13.7;Summary;171
13.8;References;171
14;Conclusion;172
14.1;References;174
15;Index;175