Voice

Speech Enhancement

• Measuring Speech Quality
  Subjective listening tests are preferred, but they are not always feasible and objective tests that attempt to predict how users will evaluate the speech quality is required.
• Echo Reduction in Coding
  Transform coding in audio and voice can introduce artifacts called pre/post echoes.
• The Cause of Echoes in Coding
  Transform coding in audio and voice can introduce artifacts called pre/post echoes.
• Battlefield Voice Activated Transmission (VOX)
  Voice Activated Transmission is commonly used in radio communications in which push-to-talk is either inconvenient or not practical, thus hands-free communication is required.
• The Applications of Voice Activity Detection (VAD)
  Voice Activity Detectors (VAD) play a major role in the telecommunications and speech processing applications.
• Speech Enhancement and Speech Intelligibility
  Speech enhancement may improve the perceptual speech quality, it does not guarantee an improvement in speech intelligibility.
• Echo Cancellation Design
  Echo canceller design considerations and related articles.
• Standard Methods of Voice Activity Detection (VAD)
  Voice Activity Detectors (VAD) play a major role in the telecommunications and speech processing applications.
• De-reverberation of Speech Signals
  De-reverberation attempts to model the impulse response of the acoustic environment and
  filter it from the received microphone signal.
• Dual Channel Noise Estimation for Speech Enhancement
  This dual-channel noise estimation technique is an extension of the single-channel noise estimation that uses the phase difference information.
• Windowing in Vocoders to Remove Artifacts
  A discussion of the use of Windowing functions to smooth discontinuities and remove audio artifacts.
• Yule-Walker Algorithm and Method Used in Voice Enhancement
  A discussion of the Yule-Walker algorithm and its use in voice enhancement applications.

Speech Recognition

• Automatic Speech Recognition and Speech Enhancement
  Discussion on the challenges of using ASR to convert speech to text.
• Forensic Speaker Recognition Preprocessor
  Discussion on using speech preprocessing to improve speech recognition rates in ASR and other applications.
• Speech Composition for Recognition
  How decomposing speech signals can help to determine the words being spoken.
• Machine Learning Algorithms
  Discussion on Machine Learning Algorithms used in advanced signal processing models.
• Measuring Detector Performance in Voice Activity Detection
  Discussion of methods for measuring VAD performance in speech enhancement systems.
• Room Quality Modeling
  Discussion on designing an acoustic space and the subjective phenomena that can be quantified for objective measurement.
• Speech Composition for Recognition
  Discussion on time and frequency characteristics of speech that can be used to determine what words are being spoken.
• Microphone Array Aided Distant Speech Recognition
  Discussion on using microphone arrays with DSR systems to improve speech recognition performance.

Speech Coders

• Comfort Noise Generation in Vocoders
  A description of Comfort Noise Generation (CNG) in Speech Coding.
• Packet Loss Concealment (PLC) in Vocoders
  A description of Packet Loss Concealment in Vocoders.
• Arithmetic Coding in Vocoders
  A description of Arithmetic Coding as it applies to Voice Coding.
• Allpass Filters in Vocoders
  A discussion of the use of all-pass filters  in Vocoders.
• Stability of LPC Filters in Vocoders
  A description of algorithms used to ensure stability of Linear Predictive Coding (LPC) filters in Speech Coders.
• Wideband Energy Detection in Scalable Speech Coding
  An algorithm is presented for determining if speech sampled at Wideband rates should be downsampled to narrowband rates.
• Efficient Implementation of LPC Analysis Filters
  An efficient algorithm for performing a LPC analysis filter is presented.
• G.729 and Its Application in the Cellular Market
  Advantages of G.729 codec for cellular applications are discussed.

Beamforming

• Acoustic Beamformer Design
  A discussion of acoustic beamformer design considerations and applications.
• Sound Propagation Models for Beamforming
  Beamforming can be applied to acoustic signal processing for speech enhancement and noise reduction.
• Differential Microphone Arrays
  Microphone arrays can be used for localization of a desired speaker/signal, tracking of the signal in the environment, and, with advanced signal processing techniques, improving the overall sound quality of the system.
• Griffiths-Jim Beamformer for Noise Reduction
  A discussion of GJB using VAD for controlled adaption of phase alignment and noise reduction of desired speech signal.
• Generation of Mixing Functions for BSS and Beamforming
  When developing acoustic noise reduction algorithms based on Beamforming and BSS, numerical simulations can be used to select the best microphone configuration for a given application.
• Speech Separation with Microphone Array
  A discussion on using beamforming and blind signal separation microphone array techniques to separate speech.
• Speaker Diarization
  Discussion on source diarization in audio video conferencing applications.
• Signal Restoration in Frequency Domain via Prediction
  A discussion on signal restoration or de-noising to reconstruct speech.
• Acoustic Two-Channel Crosstalk Canceller
  A discussion of using microphone arrays to extract signals from two acoustic sources.
• Acoustic Multiple-Channel Crosstalk Canceller
  A discussion of using microphone arrays to extract signals from multiple acoustic sources.

Noise Reduction

• Model-Based Speech Enhancement
  Maintaining harmonic information will improve the overall speech quality when compared to standard noise reduction techniques.
• Bandwidth Extension
  A discussion of algorithms and methods used to convert NB speech to WB.
• Blind Source Separation for Noise Reduction in Mobile
  Blind Source Separation or Independent Component Analysis is a multi-channel technique quite different from beamforming that can be used for noise reduction/cancellation in Mobile Telecommunications, such as cell phones and tablets
• Psychoacoustic Noise Suppression
  Psychoacoustic noise suppression takes into account the physiological and acoustic properties of the human hearing organ into the design of the algorithm for improving the perceptual quality of the communication.
• Noise Reduction in Sound Capture
  An acoustic echo canceller may need to address different acoustic and noise control environment issues. Constraints include the volume of the enclosure, the required bandwidth, and the tolerable delay.
• Noise Reduction of Non-stationary Noise Sources
  A discussion on Noise Reduction techniques in noisy environments with dynamic noise sources.
• Single Channel Noise Detection
  Methods to detect wind noise captured by microphones when used outdoors.
• Wind Noise in Mobile Telecommunication
  Discussion of algorithms to filter wind noise in hearing aids, cell phones and other mobile devices
• Monaural Speech Separation
  Speech segregation is the separation of a desired speech signal from a mix of environmental signals.
• The Artifacts of Spectral Subtraction
  An example of the artifacts of the spectral subtraction is musical noise. Musical noise are little islands of spectrum power in a signal, that appear randomly in different frequency buckets from frame to frame.
• Model-Based Speech Enhancement
  Maintaining harmonic information will improve the overall speech quality when compared to standard noise reduction techniques.
• Maintaining the Harmonic Structures for Speech Enhancement
  Maintaining harmonic information will improve the overall speech quality when compared to standard noise reduction techniques.

Particle Swarm Optimization

• Particle Swarm Optimization (PSO) in Speech Enhancement
  An application of swarm coding to two-channel noise reduction.
• Particle Swarm Optimization in Acoustic Echo Cancellation
  Particle Swarm Optimization is used to determine the optimal IIR filter for Acoustic Echo Cancellation.
• Noise Reduction using Singular Value Decomposition (SVD) and Particle Swarm Optimization (PSO)
  A method for reducing noise in speech signals based upon optimizing the effects of a Singular Value Decomposition.