Dorel Picovici

New output-based perceptual measure for predicting subjective quality of speech

The paper proposes a new output-based system for prediction of subjective speech quality, and evaluates its performance. The system is based on computing objective distance measures, such as the median minimum distance, between perceptually-based parameter vectors representing the voiced parts of the speech signal and appropriately matched reference vectors extracted from a pre-formulated codebook. The distance measures are then mapped into equivalent mean opinion scores (MOS) using regression. The codebook of the system is formed by optimally clustering the large number of speech parameter vectors extracted from an undistorted source speech database. The required clustering and matching processes are achieved by using an efficient data mining technique known as the self-organising map. The perceptually-based speech parameters are derived using perceptual linear prediction (PLP) and bark spectrum analyses. Reported evaluation results show that the proposed system is robust against speaker, utterance and distortion variations.

Output-based objective speech quality measure using self-organizing map

This paper proposes a new output-based method for assessing speech quality and evaluates its performance. The measure is based on comparing the output speech to an artificial reference signal representing the closest match from an appropriately formulated codebook. The codebook holds a number of optimally clustered speech parameter vectors, extracted from an undistorted clean speech database, and provides a reference for computing objective auditory distance measures for distorted speech. The median minimum distance is used as a measure of the objective auditory distance. The required clustering and matching processes are achieved by using an efficient data mining technique known as the self-organising map. Speech parameters derived from perceptual linear prediction (PLP) and bark spectrum analysis are used to provide speaker independent information as required by an output-based objective approach for speech quality measure.

Exploring power saving in 802.11 VoIP wireless links

Batteries are a primary resource in wireless networks for many mobile devices. Minimizing energy consumption without affecting communication activities is crucial to prolong the lifetime and improve the robustness of the wireless connection. One effective way to reduce energy consumption is to set the entire, or part of the system in a low power mode whenever possible. This paper explores a power saving approach for VoIP and presents a new power saving scheme that does not require extra protocol overhead or modifications to operate. Adopting this scheme potentially allows considerable power savings without any adverse effect on VoIP quality. Evidence of this is in simulation results, showing that a device can enter sleep mode for over 75% of the time, with no decrease in throughput.

New single-ended objective measure for non-intrusive speech quality evaluation

This article proposes a new output-based method for non-intrusive assessment of speech quality of voice communication systems and evaluates its performance. The method requires access to the processed (degraded) speech only, and is based on measuring perception-motivated objective auditory distances between the voiced parts of the output speech to appropriately matching references extracted from a pre-formulated codebook. The codebook is formed by optimally clustering a large number of parametric speech vectors extracted from a database of clean speech records. The auditory distances are then mapped into objective Mean Opinion listening quality scores. An efficient data-mining tool known as the self-organizing map (SOM) achieves the required clustering and mapping/reference matching processes. In order to obtain a perception-based, speaker-independent parametric representation of the speech, three domain transformation techniques have been investigated. The first technique is based on a perceptual linear prediction (PLP) model, the second utilises a bark spectrum (BS) analysis and the third utilises mel-frequency cepstrum coefficients (MFCC). Reported evaluation results show that the proposed method provides high correlation with subjective listening quality scores, yielding accuracy similar to that of the ITU-T P.563 while maintaining a relatively low computational complexity. Results also demonstrate that the method outperforms the PESQ in a number of distortion conditions, such as those of speech degraded by channel impairments.

Real-Time Non-Intrusive VoIP Evaluation Using Second Generation Network Processor

In todays telecommunications market there is a clear trend to adopt a flexible infrastructure to minimize costs and facilitate the introduction of new services. Typical examples are represented by dynamically managed packet-based next generation IP networks (NGN) and 4G mobile networks. In general the rate of success of these new networks will depend on whether they can deliver at least the same quality of service (QoS) than the networks they replace at lower overall cost. Voice quality represents a major dimension of perceived QoS. Measures that predict voice quality are essential in monitoring and managing the performance of such networks. Traditionally the perceived quality of voice has been evaluated by expensive and time-consuming subjective listening tests. Several attempts have been made to supplement subjective tests with objective estimators of voice quality. From the viewpoint of the measurement procedure, objective voice quality estimation methodologies can be categorized either as speech-layer objective models or packet-layer objective models

Subjective-based quality assessment for online games

This paper describes a new game assessment metric for the online gamer. The metric is in line with a mathematical model currently used for network planning assessment. In addition to the traditional network-based parameters such as delay, jitter and packet loss, new parameters based on subjective assessment are introduced. The metric aims to estimate game quality as perceived by an online game player. In order to validate and calibrate the proposed metric a subjective game quality assessment is also proposed. Two 5-point scales are introduced: a game-quality scale and a game playing-effort scale. The mean average of each scales termed, as Mean Opinion Score (MOS), will indicate the game quality (MOSGQE) and the playing-effort required (MOSGPE). Reported evaluation results indicate a high level of correlation when compared with other algorithms. Comparative results have been carried out for three online games.

Time-varying quality estimation for VoIP over Wireless Networks

Time-varying distortions are perceived as being different from stationary distortions. In terms of wireless networks the most possible time-varying distortions are based on packet loss. Such distortions degrade the VoIP quality and are directly linked to the packet-based transmission techniques. This paper proposes a new mathematical model for time-varying VoIP quality estimation. The model is extending the ITU-T legacy E-model in order to count for the perceptual relevance of the packet loss distortions. Evaluation results for VoIP over IEEE 802.11 wireless networks indicate that the proposed model outperforms the ITU-T legacy E-model and ITU-T PESQ for the investigated testing conditions.

Investigating brain signal peaks vs electroencephalograph electrode placement using multicolour 10Hz flickering graphics stimulation for Brain-computer Interface development

This paper investigates brain signal peaks and their relationship to electroencephalograph (EEG) electrode placement elicited by various multicolour stimuli flickering graphics using Steady State Visual Evoked Potential (SSVEP) paradigm. The results from various colour flickers were compared against black and white flicker. The flickers have been designed and rendered in Adobe After Effects software. 10Hz flickers were tested using 2 computers and Emotiv EPOC headset. Signal acquisition was performed using EmotivXavierTestBench application and its analysis was carried out using MATLAB with EEGLAB. All the colour flickers produced very strong 10Hz peaks in channels 1 (AF3), 3 (F3), 12 (F4) and 14 (AF4).

Investigating colour’s effect in stimulating brain oscillations for BCI systems

This paper investigates the quality and robustness of the electroencephalograph (EEG) signals elicited by various colour combinations using Steady State Visual Evoked Potential (SSVEP). The results from various colour flickers were compared against black and white flicker. The flickers have been designed and rendered in Adobe After Effects software. 10Hz flickers were tested using 3 subjects, 2 computers and Emotiv EPOC headset. Signal acquisition was performed using EmotivXavierTestBench application and its analysis was carried out using MATLAB with EEGLAB. All the colour flickers produced very strong and useable SSVEP signals.

Investigating the robustness of constant and variable period graphics in eliciting steady state visual evoked potential signals using Emotiv EPOC, MATLAB, and Adobe after effects

This paper investigates the robustness of the EEG (electroencephalograph) signal produced by SSVEP (Steady State Visual Evoked Potential) method with constant and variable period visual graphics generated in Adobe After Effects motion graphics application. The investigation has been carried out in 8 test cases using 3 subjects, 2 computers and Emotiv EPOC headset. Several external factors such as impact of ambient noises and fluorescent lighting were also considered. Results indicated that both constant and variable period flickering graphics produced equally strong EEG amplitudes. However, constant period graphics elicited cleaner signals with more pronounced base frequencies while variable period graphics exhibited strong harmonics and sub-harmonics in their signals.