AVSA Fundamentals
Various VSA softwares offer a graphic representation derived from that obtained on the Polygraph. Alpes Software uses sound recordings made in WAV format, to 11,025 Hz, 8bits. The sound is filtered by retaining low frequencies. The envelope is not the most interesting and the modulation is added in order to bring out the low frequency oscillations.
Some VSA softwares use a theory introduced by the British physiologist Olaf Lippold, 1957. They extract the low frequency of about 8 to 12 Hz and deduce a value of stress. Understandably, it's not too difficult to compute the frequency for the curve displayed Image 1, but it becomes much more complex for a curve in image 2.
Alpes Software, which is the basis and the software designer of Diogenes Lantern Pro now uses a new technique for calculating the value of stress, and not based on a direct calculation of the micro-tremor
Alpes Software uses a knowledge base collected from several VSA experts. Thousands of samples were processed, with approximately 1,200 being retained.
A totally independent scoring exercise was conducted by VSA experts. At the culmination of the project it was notable that 80% of the results had only 5% difference on the estimated values of stress, which provides a consistent and reliable basis for the techniques used by Alpes Software.
The technique consists in training a neural network with 1000 phonemes (utterances) from interrogations. The voice samples are processed to obtain the curves as the Image 1 and Image 2.
These graphs were digitized to provide a basic set to training the neural network. The training of neural network generates a file of weights that are at the heart of the network processor. Some parameters are used to adjust the neural network ; ie : the number of neurons in input, the threshold, the desired error (1%). The duration of training is approximately 45 minutes on a standard PC.
A fundamental property of the neural network is it’s capability to produce a reliable result when we enter a voice graph for which it has not been trained. When we submit a graph for which it was trained, it will produce a result with the specified error (1%).
The software AVSA extracts in real time phonemes, digitizes curves, charts are submitted to the network, then displays the results.
Another interesting neural network property is that there is no need to formalize mathematically complicated formulas to extract frequencies of micro-tremors. The neural processor teaches itself how to deduct the outcome on the basis of data provided.
It is totally free from the assumption 8Hz for low stress and 12 HZ for the presence of stress. Our only certainty is the know-how of human experts who have conducted thousands of successful VSA examinations, together with known solution outcomes.
The AVSA software can calculate up to 20 estimates of stress per second on a standard PC. Chart tests (validated protocols) are scored within 2 seconds. Exam scores (average of the tests conducted) are displayed instantly at the culmination of the last test chart.
Version 1,8A : For structured DOD interview examinations + all auto features. Manual operation is
still retained.
| Studies Validating Voice Stress Analysis (A Partial List) - Air Force Research Laboratory, Rome, NY, October, 2000. Funded by the National Institute of Justice, a three-year study by the AFRL determined that voice stress analysis achieved an accuracy rate of 100% when used to detect stress in 45 known-conclusion responses. (Available from http://extraafrl.af.mil/news/fa1100/features/detects:stress:feature.pdf)
- Cestaro, V. Department of Defense Polygraph Institute, Ft. McClellan, AL. “A Comparison Between Decision Accuracy Rates Obtained Using the Polygraph Instrument and the Computer Voice Stress Analyzer in the Absence of Jeopardy”, August, 1995. Cestaro reports that “the lab simulations established that the CVSA performs electrically according to the manufacturer’s theory of operation” and, even in the absence of jeopardy, which is a basic requirement in detection of deception, “These data indicate that there may be a systematic and predictable relationship between voice patterns and stress related to deception” (Available from DoD web site: www.dodpi.army.mil/research/research.htm)
- Tippett, R. Florida Department of Law Enforcement. “Comparative Analysis Study of the Computer Voice Stress Analyzer and Polygraph”, August, 1994. Both a polygraph and voice stress examiner, S/A Tippett examined 54 individuals that were convicted sex offenders on probation and in treatment for their crimes. His conclusions were: “With these 54 examinations, there was a 100% agreement between the CVSA and the polygraph. The number of examinees that were found to be deceptive (DI) were 35 and the number of examinees found to be not deceptive (NDI) were 19. As a result of this study, it appears that the CVSA is as effective as polygraph, which is the question this study set out to answer” (Available from University of Missouri web site: www.umr.edu/~police/cvsa/cvsamenu/htm)
- Ruiz, Legros, & Guell, 1990. Voice analysis to predict the psychological or physical state of a speaker. Published in Aviation, Space, and Environmental Medicine, 1990. Ruiz et al. reports that their “research suggests that psychological stress may be detected as acoustic modifications in the fundamental frequency of a speakers voice” and “that the fundamental frequency of the vocal signal is slowly modulated (8-14 Hz) during speech in an emotionally neutral situation. In situations demanding increased ‘mental or psychomotor’ activity, the 8-14 Hz modulation then decreases as the striated muscles surrounding the vocal cords contract in response to the arousal, thus limiting the natural trembling” (Available from Library of Congress).
- Chapman, J. Criminal Justice Department, Corning Community College, NY. “The Psychological Stress Evaluator As A Tool For Eliciting Confessions”, 1989. Chapman selected 211 criminal responses at random from 2,109 known-conclusion responses where voice stress analysis was used to test suspects. Professor Chapman’s study confirmed that voice stress analysis was accurate when utilized as a truth verification device and produced a confession rate of 94.8% of the responses where deception was indicated (Available from NITV).
- Brockway, B.F., University of Colorado School of Nursing, Denver, Colorado. “Situational Stress and Temporal Changes In Self-Report and Vocal Measurements.” Presented to the annual meeting of the American Association For the Advancement of Science, February, 1977. Brockway’s study reports that voice stress analysis does depict predictable and self-reported anxiety (Available from Library of Congress).
- Smith, G.A. “Voice Analysis For Measurement Of Anxiety.” British Journal of Medical Psychology, 1977. The author concludes that voice stress analysis is a valid measure of anxiety (Available from Library of Congress).
- Borgen, L.A., Goodman, L.I., Parke-Davis Research Laboratories, Ann Arbor, MI. “Voice Stress Analysis of Anxiolytic Drug Effects.” Results of the study indicated that voice stress analysis of the verbal responses correlated well with the other physiological responses to acute stress (Available from Library of Congress).
- Inbar, G.F., Eden, G. Dept. of Electrical Engineering Technion, Israel Institute of Technology, Haifa, Israel. “Psychological Stress Evaluators: EMG Correlation With Voice Tremor” published in Biology of Cybernetics, 1976. Inbar and Eden were able to independently verify the existence of the 8-14 Hertz ‘micro-tremor’ and to trace its origins to the central nervous system (Available from Library of Congress).
- Wiggins, S.L., McCranie, M.L., and Bailey, P. Department of Psychiatry, Medical College of Georgia, Augusta, Georgia. “Assessment of Voice Stress In Children”. Published in the Journal of Nervous Mental Disorders, 1975. The authors concluded that “audio stress can be detected with a voice stress analyzer in psychiatric patients during the course of therapy and that the VSA could serve as a useful tool for this purpose” (Available from Library of Congress).
- Heisse, J. “Is The Micro-Tremor Usable? - The Micro-Muscle Tremor In The Voice.” U.S. House Subcommittee of the Committee on Government Operations, 1974. Heisse analyzed 91 known-conclusion criminal responses utilizing voice stress analysis and determined that “Audio stress analysis seems to be valid in detecting changes in various psycho physiological parameters so that a trained examiner utilizing standard techniques can evaluate these changes and thus utilize the instrument in truth and deception” (Available from Library of Congress).
- Brenner, M. “Stage Fright and Steven’s Law.” Dept. of Psychology, Ohio State University, presented at the convention of the Eastern Psychological Association, April, 1974. Brenner, utilizing a voice stress analyzer, established that the frequency of vocal stress increased as a function of audience size (Available from Library of Congress).
- Lippold, O. “Oscillations In The Stretch Reflex Arc And The Origin Of The Rhythmical 8-12 C/S Component Of The Physiological Tremor.” The Journal Of Physiology, February, 1970. Lippold first discovers the physiological tremor in the human voice in the 8-12 Hz range (Available from Library of Congress).
- Lippold, O., Redfearn, J., Vuco, R. “The Rhythmical Activity Of Groups Of Muscle Units In The Voluntary Contraction Of Muscle.” The Journal Of Physiology, August, 1957. Lippold, Redfearn and Vuco begin exploring the correlation between muscle activity and stress (Available from Library of Congress).
|
Studies of Voice Stress Analysis (External Links)
|
