THEORY EXAMINATION (SEM–VIII) 2016-17 SPEECH PROCESSING

SECTION A – Fundamental Concepts of Speech Processing

Section A contains short conceptual questions designed to test the basic understanding of speech signals, digital signal processing, and speech analysis techniques. 

Question (a): What is Pitch? Explain.

Answer:
Pitch refers to the perceptual property of sound that determines whether a sound is perceived as high or low by the human ear. In speech processing, pitch corresponds to the fundamental frequency of the speech signal produced by the vibration of the vocal cords.

When the vocal cords vibrate quickly, the pitch is high, and when they vibrate slowly, the pitch becomes low. Pitch plays an important role in speech processing systems because it helps in identifying the speaker and distinguishing between voiced and unvoiced speech sounds.

Pitch detection is commonly used in applications such as speech recognition, speaker identification, and speech synthesis.

Question (b): Explain Acoustic Phonetics.

Answer:
Acoustic phonetics is the branch of phonetics that studies the physical properties of speech sounds. It focuses on how speech signals are produced, transmitted through the air, and perceived by the human ear.

Acoustic phonetics examines properties such as frequency, amplitude, duration, and spectral characteristics of speech signals. By analyzing these properties, researchers can understand how different speech sounds are formed and how they can be processed digitally.

This field is important in developing technologies such as speech recognition systems and speech synthesis systems.

Question (c): Why is Sampling Required?

Answer:
Sampling is required to convert an analog speech signal into a digital signal so that it can be processed by digital systems such as computers.

Speech signals are continuous in nature. However, digital systems operate using discrete values. Sampling captures the amplitude of the signal at regular intervals to represent the continuous signal digitally.

According to the Nyquist theorem, the sampling frequency must be at least twice the maximum frequency present in the signal to avoid distortion. For example, telephone speech signals are typically sampled at 8 kHz.

Sampling enables digital storage, transmission, and processing of speech signals.

Question (d): Define Channel Vocoder.

Answer:
A channel vocoder is a speech processing system used for speech analysis, compression, and synthesis. It divides the speech signal into several frequency channels using band-pass filters.

Each channel analyzes the energy present in a specific frequency band. The system then extracts important parameters such as amplitude and pitch instead of transmitting the entire speech waveform.

By transmitting only these parameters, the vocoder significantly reduces the amount of data required for speech communication.

Channel vocoders are commonly used in telecommunications and speech compression systems.

Question (e): What is Frequency Domain?

Answer:
The frequency domain represents a signal in terms of its frequency components rather than time.

In speech processing, analyzing signals in the frequency domain helps identify characteristics such as pitch, harmonics, and formants. This representation makes it easier to analyze how different frequencies contribute to the overall speech signal.

Techniques such as the Fourier Transform are used to convert signals from the time domain into the frequency domain.

Question (f): Define Correlation Function with Example.

Answer:
The correlation function measures the similarity between two signals or between a signal and a delayed version of itself.

In speech processing, correlation functions are used for tasks such as pitch detection and pattern recognition.

For example, when a speech signal repeats periodically, the correlation function produces peaks at intervals corresponding to the pitch period. This helps determine the fundamental frequency of the speech signal.

Question (g): What is a Filter? Explain.

Answer:
A filter is a device or algorithm used to modify a signal by removing unwanted components or enhancing specific frequency components.

In speech processing, filters are used to eliminate background noise and isolate important frequency bands of speech signals.

Common types of filters include:

Low-pass filters

High-pass filters

Band-pass filters

Band-stop filters

Filters improve the clarity and quality of speech signals.

Question (h): Differentiate Between Speech and Silence.

Speech processing systems detect silence segments to improve efficiency and reduce unnecessary processing.

Question (i): Define Convolution with Example.

Answer:
Convolution is a mathematical operation used to combine two signals to produce a third signal.

In speech processing, convolution is used to model how speech signals pass through systems such as filters or the vocal tract.

For example, when a speech signal passes through a filter, the output signal is the convolution of the input signal and the filter's impulse response.

Convolution is widely used in digital signal processing for system analysis.

Question (j): What is Linear Predictive Coding (LPC)?

Answer:
Linear Predictive Coding is a method used in speech processing to represent speech signals efficiently.

LPC predicts the current speech sample based on a linear combination of previous speech samples. It extracts parameters that represent the spectral envelope of the speech signal.

LPC is widely used in applications such as speech synthesis, speech compression, and voice communication systems.

SECTION B – Intermediate Concepts of Speech Processing

Section B focuses on speech signal modeling, parameter extraction, and speech analysis techniques. 

Question: Sampling and Quantization in Speech Signals

Sampling and quantization are two processes used to convert analog speech signals into digital form.

Sampling captures the amplitude of the signal at regular intervals. Quantization converts the sampled amplitudes into discrete numerical levels that can be stored digitally.

For example, when recording speech using a microphone, the analog signal is sampled and quantized before being stored as digital audio.

These processes enable digital speech processing and storage.

Question: Digital Models for Speech Signals

Digital models represent speech signals mathematically to help analyze and synthesize speech.

One common model is the source-filter model, which assumes that speech production involves a sound source (vocal cords) and a filter (vocal tract).

The vocal tract shapes the sound produced by the vocal cords to create different speech sounds.

This model is widely used in speech synthesis systems.

Question: Applications of Speech Processing

Speech processing has many practical applications, including:

Speech recognition systems

Voice assistants

Speaker identification

Automated customer service systems

Hearing aids

Voice-controlled devices

These technologies improve communication between humans and computers.

Question: Short-Term Pitch Detection

Short-term pitch detection estimates the pitch of speech signals within short time frames.

The process involves dividing the speech signal into short frames, computing correlation functions, and identifying peaks corresponding to pitch periods.

Pitch detection helps determine whether speech is voiced or unvoiced.

SECTION C – Advanced Concepts of Speech Processing

Section C focuses on advanced techniques such as speech synthesis, Fourier analysis, and speech parameter estimation. 

Question: Speech Synthesis and LPC

Speech synthesis refers to generating artificial speech using computers.

Speech synthesis systems convert text or symbolic information into speech signals. These systems typically involve stages such as text analysis, phoneme generation, and waveform synthesis.

Linear Predictive Coding plays a significant role in speech synthesis because it models the vocal tract and generates realistic speech signals.

LPC uses mathematical equations to estimate predictor coefficients that describe speech signals efficiently.

Question: Short-Time Fourier Analysis

Short-Time Fourier Transform (STFT) is used to analyze how the frequency components of speech signals change over time.

Speech signals are non-stationary, meaning their properties vary over time. STFT divides the signal into small time frames and computes the Fourier transform for each frame.

This allows visualization of speech signals using spectrograms, which display frequency variations over time.

Question: Autocorrelation, NMSE, and Formant Estimation

Autocorrelation Method
Autocorrelation measures similarity between a signal and delayed versions of itself. It is widely used for pitch detection.

Normalized Mean Square Error (NMSE)
NMSE measures the difference between predicted and actual speech signals. It is used to evaluate the accuracy of speech models.

Formant Estimation
Formants are resonance frequencies of the vocal tract. They help identify vowel sounds and play an important role in speech recognition systems.

Conclusion

Speech processing is an interdisciplinary field that combines signal processing, linguistics, and computer science to analyze and synthesize speech signals. Concepts such as pitch detection, sampling, filtering, and linear predictive coding are fundamental to modern speech technologies.

These techniques enable applications such as voice assistants, speech recognition systems, and speech synthesis technologies that are widely used in modern communication systems.