DSP_FOEHU - Lec 13 - Digital Signal Processing Applications I
3 likes2,172 views
This document provides an overview of digital signal processing applications including digital spectrum analysis, speech processing, and radar. It discusses different types of digital spectrum analyzers including filter bank, swept, and FFT analyzers. It also covers topics related to speech processing like the anatomy of speech production, speech perception, voiced and unvoiced sounds, and phonemes. Common speech coding techniques are introduced such as vocoding, ADPCM, LPC, and CELP coding. Radar applications of DSP are also briefly mentioned.
DSP_FOEHU - Lec 13 - Digital Signal Processing Applications I
- 1. رَـدْقِـن،،،لمااننا نصدقْْقِنرَد LECTURE (13) Digital Signal Processing Applications (I) Assist. Prof. Amr E. Mohamed
- 2. Agenda Introduction Digital Spectrum Analyzer Speech Processing Radar 2
- 3. Introduction Three levels of understanding are required to appreciate the technological advancement of DSP Theoretical Conceptual Practical Practice Concepts Theory Ability to implement theory and concepts in working code (MATLAB, C, C++, JAVA) Basic understanding of how theory is applied Mathematics, derivations, signal processing 3
- 5. Topics Time Domain & Frequency Domain. Fourier Analysis. Decibel. Spectrum Analyzer. Filter Bank Spectrum Analyzer. Swept Spectrum Analyzer. FFT Spectrum Analyzer. Real-Time Spectrum Analyzer. 5
- 6. Time and Frequency Measurements 6
- 7. Time and Frequency Measurements A time domain graph shows how the signal amplitude changes over time. To fully understand the performance of your device/system, you will also want to analyze the signal(s) in the frequency-domain. The spectrum analyzer is to the frequency domain as the oscilloscope is to the time domain. 7
- 8. Time and Frequency Measurements Some fields of application Telecommunications. Analog and Power Electronics. Electromagnetic. Electrical Systems and Machines Bioengineering 8
- 9. Frequency Response Fourier analysis: Fourier Series. Fourier Transform. Discrete-Time Fourier Transform (DTFT). Discrete Fourier Transform (DFT). 9
- 11. IMPORTANCE AND USAGE OF DECIBEL 11
- 12. IMPORTANCE AND USAGE OF DECIBEL 12
- 13. IMPORTANCE AND USAGE OF DECIBEL 13
- 14. IMPORTANCE AND USAGE OF DECIBEL 14
- 15. IMPORTANCE AND USAGE OF DECIBEL 15
- 16. IMPORTANCE AND USAGE OF DECIBEL 16
- 18. 1. Bank-of-filters Spectrum Analyzer 18
- 19. 1. Bank-of-filters Spectrum Analyzer 19
- 20. 1. Bank-of-filters Spectrum Analyzer Main Design Fixed pass-band filters Minimum overlap between the frequency responses of two adjacent filters. Detectors for measuring the ”sinusoidal” output of the filters. Frequency resolution It is strictly dependent on RBW (resolution bandwidth) value Drawbacks Example: 20
- 21. 2. Swept Spectrum Analyzers Sequential Spectrum Analyzer. Variable tuning filter spectrum analyzer. 21
- 22. 2. Swept Spectrum Analyzers 22
- 23. Main design choices Single pass-band filter with constant central frequency Mixer implementing the superheterodyne technique Detector for measuring the envelope of the filter output signal 23 3. Swept Superheterodyne Spectrum Analyzers
- 24. 24 3. Swept Superheterodyne Spectrum Analyzers
- 25. 25 3. Swept Superheterodyne Spectrum Analyzers
- 26. 4. FFT spectrum analyzer 26
- 28. Speech Generation and Perception The study of the anatomy of the organs of speech is required as a background for articulatory and acoustic phonetics. An understanding of hearing and perception is needed in the field of both speech synthesis and speech enhancement and is useful in the field of automatic speech recognition. 28
- 29. 29 Schematic diagram of the human speech production
- 30. Organs of Speech Lungs and trachea: source of air during speech. The vocal organs work by using compressed air; this is supplied by the lungs and delivered to the system by way of the trachea. These organs also control the loudness of the resulting speech. The Larynx: This is a complicated system of cartilages and muscle containing and controlling the vocal cords. The vocal cords are composed of twin infoldings of mucous membrane stretched horizontally, from back to front, across the larynx. They vibrate, modulating the flow of air being expelled from the lungs during phonation. Open when breathing and vibrating for speech or singing, the folds are controlled via the vagus nerve. 30
- 31. Organs of Speech The Vocal Tract: The vocal tract is the cavity in human beings and in animals where sound that is produced at the sound source is filtered. • Oral cavity: Forward of the velum and bounded by lips, tongue and palate • Nasal cavity: Above the palate and extending from the pharynx to the nostrils 31
- 32. A General Discrete-Time Model For Speech Production The operation of the system is divided into two functions : Excitation Modulation Excitation (glottis) Modulation (vocal tract) Radiate speech 32
- 33. Hearing and perception Hearing is a process which sound is received and convert into nerve impulse Perception is the post-processing within the brain by which the sounds heard are interpreted and given meaning 33
- 34. What is the difference between consonant and vowel? A consonant is a sound that is made with the air stopping once or more during the vocalization. That means that at some point, the sound is stopped by your teeth, tongue, lips, or constriction of the vocal cords. Voiced Unvoiced A vowel is a sound that is made with the mouth and throat not closing at any point. Tongue high or low Tongue front or back Lips rounded or unrounded 34
- 35. Voiced and Unvoiced sounds There are two mutually exclusive ways excitation functions to model voiced and unvoiced speech sounds. For a short time-basis analysis: voiced speech is considered periodic with a fundamental frequency of 𝑓0, and a pitch period of 1/𝑓0, which depends on the speaker. Hence, Voiced speech is generated by exciting the all pole filter model by a periodic impulse train. On the other hand, unvoiced sounds are generated by exciting the all- pole filter by the output of a random noise generator 35
- 36. Voiced/Unvoiced The fundamental difference between these two types of speech sounds comes from the following: the way they are produced. The vibrations of the vocal cords produce voiced sounds. The rate at which the vocal cords vibrate dictates the pitch of the sound. On the other hand, unvoiced sounds do not rely on the vibration of the vocal cords. The unvoiced sounds are created by the constriction of the vocal tract. The vocal cords remain open and the constrictions of the vocal tract force air out to produce the unvoiced sounds Given a short segment of a speech signal, lets say about 20𝑚𝑠 or 160 samples at a sampling rate 8𝐾𝐻𝑧, the speech encoder at the transmitter must determine the proper excitation function, the pitch period for voiced speech, the gain, and the coefficients 36
- 37. Phonemics (phonemes): A phoneme is the smallest sound unit in a given language that is sufficient to differentiate one word from another English Phonemes Vowels Semi-vowels Fricatives Nasals Stops Aspiration uw ux uh ah ax ah-h aa ao ae eh ih ix ey iy ay ow aw oy er axr el y r l el w jh ch s sh z zh f th v dh m n ng em en eng nx b d g p t k dx q bcl dcl gcl pcl tcl kcl hv hh 37
- 38. 1. The Channel Vocoder Application 38
- 39. The Channel Vocoder - analyzer The channel vocoder employs a bank of bandpass filters, Each having a bandwidth between 100HZ and 300HZ. Typically, 16-20 linear phase FIR filter are used. The output of each filter is rectified and lowpass filtered. The bandwidth of the lowpass filter is selected to match the time variations in the characteristics of the vocal tract. For measurement of the spectral magnitudes, a voicing detector and a pitch estimator are included in the speech analysis. 39
- 40. The Channel Vocoder – analyzer (block diagram) Bandpass Filter A/D Converter Lowpass Filter A/D Converter Lowpass Filter Rectifier Rectifier Bandpass Filter Voicing detector Pitch detector Encoder S(n) To Channel 40
- 41. The Channel Vocoder - synthesizer At the receiver the signal samples are passed through D/A converters. The outputs of the D/As are multiplied by the voiced or unvoiced signal sources. The resulting signal are passed through bandpass filters. The outputs of the bandpass filters are summed to form the synthesized speech signal. 41
- 42. The Channel Vocoder - synthesizer (block diagram) D/A ConverterDecoder D/A Converter Voicing Information Pitch period Pulse generator Random -Noise generator Bandpass Filter Bandpass Filter Switch ∑ Output speech From Channel 42
- 43. Adaptive Differential Pulse Code Modulation The ADPCM coder takes advantage of the fact that neighboring audio samples are generally similar to each other. Instead of representing each audio sample independently as in Pulse Code Modulation (PCM), an ADPCM encoder computes the difference between each audio sample and its predicted value and outputs the PCM value of the differential. 43
- 45. Linear Predictive Coding The objective of LP analysis is to estimate parameters of an all-pole model of the vocal tract. Several methods have been devised for generating the excitation sequence for speech synthesizes. LPC-type of speech analysis and synthesis are differ primarily in the type of excitation signal that is generated for speech synthesis. 45
- 46. LPC 10 This methods is called LPC-10 because of 10 coefficient are typically employed. LPC-10 partitions the speech into the 180 sample frame. Pitch and voicing decision are determined by using the Average Magnitude Difference Function (AMDF) and zero-crossing measures. 46
- 47. Residual Excited LP Vocoder Speech quality in speech quality can be improved at the expense of a higher bit rate by computing and transmitting a residual error, as done in the case of DPCM. One method is that the LPC model and excitation parameters are estimated from a frame of speech. 47
- 48. Residual Excited LP Vocoder The speech is synthesized at the transmitter and subtracted from the original speech signal to form the residual error. The residual error is quantized, coded, and transmitted to the receiver At the receiver the signal is synthesized by adding the residual error to the signal generated from the model. 48
- 49. RELP Block Diagram Buffer And window LP analysis ∑ Encoder LP - Synthesis model S(n) To ChannelExcitation parameters LP - Parameters 49
- 50. Code Excited LP CELP is an analysis-by-synthesis method in which the excitation sequence is selected from a codebook of zero-mean Gaussian sequence. The bit rate of the CELP is 4800 bps. 50
- 51. CELP (analysis-by-synthesis coder) Gaussian Excitation codebook Pitch Synthesis filter Spectral Envelope (LP) Synthesis filter ∑ Perceptual Weighting Filter W(z) Compute Energy of Error (square and sum) Buffer and LP analysis Side information Gain LP - parameters Speech samples Index of Excitation sequence + - 51
- 52. CELP (synthesizer) From Channel decoder Buffer & controller Gaussian Excitation codebook Pitch Synthesis filter LP Synthesis filter LP parameters, gain and pitch estimate updates 52
- 53. 2. Text To Speech (TTS) Application A text to speech synthesizer is a computer based system that should be able to read any text whether it was directly introduced into the computer or through character recognition system (OCR). And speech should be intelligible and natural. 53
- 54. Speech Synthesis Concept Text to Phone Sequence Phone Sequence to Speech Text Speech Natural Language Processing (NLP) Speech Processing (DSP Module) Text Speech 54
- 55. NLP and DSP Modules The Natural Language Processing (NLP) module is capable of producing a phonetic transcription of the text to be read, together with the desired intonation and rhythm. It takes in the text as input and give narrow phonetic transcription as output which is further forwarded to the DSP module. the DSP module which transforms the symbolic information it receives into natural sounding speech. “Narrow phonetic transcription” which is taken as intermediate varies from synthesizer system to another. 55
- 56. 3- Speech enhancement Wiener Filtering: A linear estimation of clean signal from the noisy signal Using MMSE criterion tttttt ttt vyaazzyEy vyz ˆ noiseadditiveFor 1KSpeechNoisy 1KNoise 1KSpeechClean t t t z v y 56
- 57. Wiener Filter zz vvzz vvyy yy vvyyzz S SS jH SS S jH SSS 57
- 60. Radar Radar: RAdio Detection And Ranging Need a directional radio beam. Measure time between transmit pulse and receive pulse Find Distance: Divide speed of light by interval time 60
- 61. Doppler concept – frequency shift through motion 61
- 62. Doppler effect Frequency shift in received pulse: fDoppler = 2 vrelative / λ Example: assume X band radar operating at 10GHz (3cm wavelength) Airborne radar traveling at 500 mph Target 1 traveling away from radar at 800 mph Vrelative = 500 – 800 = -300 mph = -134 meter/s Target 2 traveling towards radar at 400 mph Vrelative = 500 + 400 = 900 mph = 402 meter/s First target Doppler shift = 2 (-134m/s) / (0.03m) = - 8.93 kHz Second target Doppler shift = 2 (402m/s) / (0.03m) = 26.8 kHz 62
- 64. 64