Classification of speech under stress by physical modeling

Speech of human beings is the reflection of the state of mind. Proper evaluation of these speech signals into stress types is necessary in order to ensure that the person is in a healthy state of mind. In this work we propose a SVM classifier for speech stress classification algorithm, with sophisticated feature extraction techniques as Mel Frequency Cepstral Coefficients (MFCC). The SVM algorithm assists the system to learn the speech patterns in real time and self-train itself in order to improve the classification accuracy of the overall system. The proposed system is suitable for real time speech and is language and word independent.

Proceedings of the 1998 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP '98 (Cat. No.98CH36181)

This study proposes a new set of feature parameters based on subband analysis of the speech signal for classi cation of speech under stress. The new speech features are Scale Energy SE, Autocorrelation-Scale-Energy ACSE, Subband based cepstral parameters SC, and Autocorrelation-SC ACSC. The parameters' ability to capture di erent stress types is compared to widely used Mel-scale cepstrum based representations: Mel-frequency cepstral coe cents MFCC and Autocorrelation-Mel-scale AC-Mel. Next, a feedforward neural network is formulated for speaker-dependent stress classi cation of 10 stress conditions: Angry, Clear, Cond50 70, Fast, Loud, Lombard, Neutral, Question, Slow, and Soft. The classi cation algorithm is evaluated using a previously established stressed speech database SUSAS 4. Subband based features are shown to achieve + 7:3 and +9:1 increase in the classi cation rates over the MFCC based parameters for ungrouped and grouped stress closed vocabulary test scenarios respectively. Moreover the average scores across the simulations of new features are +8:6 and +13:6 higher than MFCC based features for the ungrouped and grouped stress test scenarious respectively.

International Journal of Bioinformatics Research and Applications, 2015

When a person is emotionally charged, stress could be discerned in his voice. This paper presents a simplified and a non-invasive approach to detect psycho-physiological stress by monitoring the acoustic modifications during a stressful conversation. Voice database consists of audio clips from eight different popular FM broadcasts wherein the host of the show vexes the subjects who are otherwise unaware of the charade. The audio clips are obtained from real-life stressful conversations (no simulated emotions). Analysis is done using PRAAT software to evaluate mean fundamental frequency (F0) and formant frequencies (F1, F2, F3, F4) both in neutral and stressed state. Results suggest that F0 increases with stress; however, formant frequency decreases with stress. Comparison of Fourier and chirp spectra of short vowel segment shows that for relaxed speech, the two spectra are similar; however, for stressed speech, they differ in the high frequency range due to increased pitch modulation.

9th International Conference on Affective Computing and Intelligent Interaction (ACII), 2021

This paper investigates a robust and effective automatic stress detection model based on human vocal features. Our study experimental dataset contains the voices of 58 Greekspeaking participants (24 male, 34 female, 26.9±4.8 years old), both in neutral and stressed conditions. We extracted a total of 76 speech-derived features after extensive study of the relevant literature. We investigated and selected the most robust features using automatic feature selection methods, comparing multiple feature ranking methods (such as RFE, mRMR, stepwise fit) to assess their pattern across gender & experimental phase factors. Then, classification was performed both for the entire dataset, and then for each experimental task, for both genders combined and separately. The performance was evaluated using 10-fold cross-validation on the speakers. Our analysis achieved a best classification accuracy of 84.8% using linear SVM for the social exposure phase and 74.5% for the mental tasks phase using the gaussian SVM classifier. The ordinal modelling improved significantly our results, yielding a best on-subject basis 10fold cross-validation classification accuracy of 95.0% for social exposure using gaussian SVM and 85.9% for mental tasks using the gaussian SVM. From our analysis, specific vocal features were identified as being robust and relevant to stress along with parameters to construct the stress model. However, it is was observed the susceptibility of speech to bias and masking and thus the need for universal speech markers for stress detection.

Lecture Notes in Computer Science

In this chapter, we consider a range of issues associated with analysis, modeling, and recognition of speech under stress. We start by defining stress, what could be perceived as stress, and how it affects the speech production system. In the discussion that follows, we explore how individuals differ in their perception of stress, and hence understand the cues associated with perceiving stress. Having considered the domains of stress, areas for speech analysis under stress, we shift to the development of algorithms to estimate, classify or distinguish different stress conditions. We will then conclude with revealing what might be in store for understanding stress, and the development of techniques to overcome the effects of stress for speech recognition and human-computer interactive systems.

2006 IEEE International Conference on Acoustics Speed and Signal Processing Proceedings

Presently, automatic stress detection methods for speech employ a binary decision approach, deciding whether the speaker is or is not under stress. Since the amount of stress a speaker is under varies and can change gradually, a reliable stress level detection scheme becomes necessary to accurately assess the condition of the speaker. Such a capability is pertinent to a number of applications, such as for those personnel in law enforcement positions. Using speech and biometric data collected from a realworld, variable-stress level law enforcement training scenario, this study illustrates two methods for automatically assessing stress levels in speech using a hybrid multi-dimensional feature space comprised of frequency-based and Teager Energy Operator-based features. The first approach uses a nearest neighbor-type clustering scheme at the vowel token level to classify speech data into one of three levels of stress, yielding an overall error rate of 50.5%. The second approach employs accumulated Euclidean distance metric weighting at the sentence-level to yield a relative improvement of 12.1% in performance.

Interspeech 2008, 2008

International Journal of Engineering & Technology, 2018

The Detection of stress from speech signal is gaining large attention recently. The emergence of new methods and techniques for feature extraction and classification paved the way to different solutions to detect different stress conditions using human speech and led to an in-crease in the accuracy of stress recognition. A large number of parameters are proposed for the characterization of stress in speech. Similarly numerous classifiers and machine learning algorithms are investigated for stress classification and regression. In this treatise, a recital on the commonly used databases, stress conditions, different feature extraction methods and classifiers along with some of the statistical measures as well as compensation techniques for stress detection are presented in this article. After thorough illustration of existing methodology for the task, future prospects for the work are elaborated.

IEEE Transactions on Speech and Audio Processing, 2000

It is well known that the performance of speech recognition algorithms degrade in the presence of adverse environments where a speaker is under stress, emotion, or Lombard effect. This study evaluates the effectiveness of traditional features in recognition of speech under stress and formulates new features which are shown to improve stressed speech recognition. The focus is on formulating robust features which are less dependent on the speaking conditions rather than applying compensation or adaptation techniques. The stressed speaking styles considered are simulated angry and loud, Lombard effect speech, and noisy actual stressed speech from the SUSAS database which is available on CD-ROM through the NATO IST/TG-01 research group and LDC 1 . In addition, this study investigates the immunity of linear prediction power spectrum and fast Fourier transform power spectrum to the presence of stress. Our results show that unlike fast Fourier transform's (FFT) immunity to noise, the linear prediction power spectrum is more immune than FFT to stress as well as to a combination of a noisy and stressful environment. Finally, the effect of various parameter processing such as fixed versus variable preemphasis, liftering, and fixed versus cepstral mean normalization are studied. Two alternative frequency partitioning methods are proposed and compared with traditional mel-frequency cepstral coefficients (MFCC) features for stressed speech recognition. It is shown that the alternate filterbank frequency partitions are more effective for recognition of speech under both simulated and actual stressed conditions.