LIE DETECTION SYSTEM USING ARTIFICIAL NEURAL NETWORK
issue 1

LIE DETECTION SYSTEM USING ARTIFICIAL NEURAL NETWORK

1.Gajal S. Agrawal, 2.Dipali R. Koche, 3.Himani D. Nagrale, 4.Prof. V.R. Surjuse
1,2,3Student, Computer Technology, K.D.K. College of Engineering, Nagpur, India
4. Assistant Professor, Computer Technology, K.D.K. College of Engineering, Nagpur, India

Abstract:

In this paper, we demonstrate that we can use non-invasive physiology sensing to detect stress and lying, within the context of Artificial Neural Network (ANN). We show how simply derived non- invasive physiological features like voice pitch variation, and pulse variability are correlated to sort of high stress situations found in world. Using these features, we will develop simple linear models which will be wont to identify stress and bluffing.

Keywords: ANN, stress, voice pitch variation, pulse.

I. INTRODUCTION

The National Institute of Occupational Safety and Health states that stress is becoming the most prevalent reason for worker disability A 1992 UN report referred to as job stress “The twentieth Century Epidemic”, whereas the world Health Organization expressed in 1996 that stress was a “World Wide Epidemic” .Researchers estimate job stress costs American industries between $200 and 300 billion annually. Given the mounting social prices of stress, the possibility of automatically identifying and monitoring stress levels for intervention purposes is compelling. This is notably true for people that often had high stress environments, such as financial traders or emergency workers. In such things, faulty performance as a result of acute stress can lead to million dollars loss of or even the loss of life. Given the implications of stress on work performance, there has been recent interest in observation the work performance of people underneath stress. Specifically, studies in behavioral finance gift physiological proof that even the foremost sea zoned securities bargainer exhibits vital emotional response as measured by elevated levels of skin electrical phenomenon and vessel variables throughout bound transient market events [Lo & Repin 2002, Lo et al. 2005]. Other studies have supported the proof linking feeling with commerce performance [Steenbarger 2002]. These studies indicate that psychophysiology and stress are intimately linked, and it is possible to infer one from the other. We wish to demonstrate that non-invasively derived physiology and behavioural sensing can be correlated with various stressful events in poker tournaments. Specifically, we’ll be viewing at subjective reports of stress levels, bluffing, and allign things as outcomes and correlating these things with physiology options mass over the hand.

II. LITERATURE REVIEW

Whereas ancient clinical physiology observance focuses on distinguishing correct physiological responses (e.g. ECG traces during a heart arrhythmia), long-term monitoring enabled by minimally invasive sensing provides the ability to correlate con-textual measures over time to a person’s behaviour and internal state. Research has shown that it’s potential to correlate minimally invasive physiology measures to identify notions of a person’s intentions and spirit like interest, happiness, and stress [Picard 2001, Picard et al. 2001]. In specific, accurately distinguishing human intention like lying has been notably contentious given its inherently subjective nature. The idea of the „lie detector‟ has invariably captured the imagination and interest of the favoured press since it had been fabricated over 100 years a gone in 1902 by James Mackenzie. The first version fabricated by Mackenzie, was additionally known as the medical instrument tool as a result of it checked out a variety of physiological phenomena like a person’s rate, respiration rate, vital sign, and skin physical phenomenon whereas someone is questioned. Though a report released in 2002 by America’s National Academy of Sciences indicated that poly-graphs are not completely reliable (better as a measure of stress than veracity), there is no doubt that a range of physiological phenomena are often correlative to lying. If such a ambiguous factor as telling the reality will be foreseen and related to basic physiological options like those analyzed by a medical instrument, it’s not a large leap of religion to believe mistreatment physiology to quantify a person’s different internal states like stress, frustration, interest level, excitement, attention, drowsiness, and have an impact on [Picard 2001, Picard et. al 2001].In fact, there’s analysis to indicate that these items will in truth be accurately quantified, and moreover that these measures will demonstrate a high degree of correlation and concordance among interacting groups of individuals , whether being stimulated while passively watching a movie [Madan et al. 2004] or the extremely concerned interaction dynamics of a medicine patient and a healer [Marci 2002]. Barbour’s article about the polygraph and the privacy tort contributed to understand cons of polygraph. The aim of the article was that showing polygraph misuses, detail information about using polygraph in law enforcement, explaining what a polygraph is, what a polygraph test includes, weaknesses of detection of truth, and validity and reliability of it. The films which are about finding a suspected and criminal person helps to accommodate the idea of identifying guilty through using polygraph. On the other hand, it did not work as seeming in the films. Sometimes, it causes many innocent people are accepted as a guilty. There was an evident about it; a woman was killed in front of her boy. After the horrible event, the boy was accepted suspected person, and he took polygraph test. Because of he was sad about his mom’s death; the polygraph assumed that he was guilty. After police officer found another evidence, they were convinced that he did not kill her mother. Moreover, people that are in doubtful condition are worried and scared about taking polygraph testing. Because they try to convince employees about, they are innocent. And their endeavor causes physiological change in their body, and the result of it, they are accepted offender. By looking all of them, the polygraph is not useful test for lie detecting. According to this paper, there is no specific physiological symptom for lying, and there is no appropriate machine to detect liars (Barbour, 2002). Ekman had done a study about specific micro expression to detect lies. Pauses and speech errors, raised pitch of voice, louder speech, sweating, blinking, faster/shallower breathing, whitening of the face are some examples of clues. There are some unreal facial expressions of emotion also. For example; happiness should involve eye muscles, otherwise it’s false, asymmetrical facial expressions occur, and fear and sadness involve a characteristic forehead expression especially involving the eyebrows. Ekman also developed 38-item lying checklist to become detecting lies in easier way (Ekman, 1996) Voice Stress Analysis was another methodology to use implying perpetrators. It was better method than polygraph. Lies are often detected through analyzing speech processes of persons. A tremor in speech give clue to the person is lying. Conversely, it can’t be assumed that objective method to lie detection. Because of many people have problem when they speak with foreign people, and they can experience a tremor in their speech. We cannot generalize it in whole events. The VSA also wasn’t wont to be as a polygraph . Lastly, catching lie is important issue in criminal or forensic area. Hence, deception is common in legal area. In history, many methods have been used to determine deceptions. Some of them could measure to physiological changes in human body when lying process occurs, but has limitation to evaluate. For example, they need to physical contact to human body and it creates uncontrollable fear and anxiety situation to people. So, they have not been accepted lie detector. On the other hand, Ekman et al found a different method without contacting with people; it is done by looking human’s behavioural and emotional expression. This method also includes some limitation, but it can be accepted more reliable than other. To sum up, legal area still needs far more reliable and valid method of lie detection system..

III.METHODOLOGY

  1. Algorithms Used:
    1. Two Class Decision Jungle Algorithm: The Two-Class Decision Jungle module returns an untrained classifier. You then train this model on a labeled training data set, by using Train Model. Decision jungles have the following advantages:
      • By allowing tree branches to merge, a decision DAG typically has a lower memory footprint and better generalization performance than a decision tree, albeit at the cost of somewhat longer training time.
      • Decision jungles are non-parametric models that can represent non-linear decision boundaries.
      • • They perform integrated feature selection and classification and are resilient in the presence of noisy features
    2. Two Class Support Vector Machine: Support vector machines (SVMs) are a well-researched class of supervised learning methods. This particular implementation is suited to prediction of two possible outcomes, supported either continuous or categorical variables. Support vector machines are among the earliest of machine learning algorithms, and SVM models are utilized in many applications, from information retrieval to text and image classification. SVMs are often used for both classification and regression tasks. This SVM model may be a supervised learning model that needs labeled data. In the training process, the algorithm analyzes input file and recognizes patterns during a multi-dimensional feature space called the hyperplane. All input examples are represented as points during this space, and are mapped to output categories in such how that categories are divided by as wide and clear a gap as possible.
    3. Polygraph Algorithm: The algorithm has been developed on polygraph tests for actual criminal cases provided by the DoDPI. The input to PolyScore is that the digitized polygraph signal, and therefore the output may be a probability of deception based either on a logistic regression or a neural network model. A critical a part of polygraph examination is that the analysis and interpretation of the physiological data recorded on polygraph charts. Currently, polygraph examiners believe their subjective global evaluation of the charts, various partly objective numerical scoring methods, computerized algorithms for chart scoring, or some combination of the three. Computerized systems have the potential to scale back bias within the reading of charts and eliminate problems of imperfect inter-rater variability that exist with human scoring. The extent to which they will improve accuracy depends on how one views the appropriateness of using other knowledge available to examiners, like demographic information, historical background of the subject, and behavioral observations.
    4. Linear Regression Algorithm: Linear Regression is a machine learning algorithm based on supervised learning. Linear regression performs the task to predict a variable value (y) supported a given experimental variable (x). So, this regression technique finds out a linear relationship between x (input) and y(output).
  1. DLIB libraries: Dlib is a general purpose cross-platform software library written in the programming language C++. Its design is heavily influenced by ideas from design by contract and component-based software engineering. Thus it is, first and foremost, a group of independent software components. It is open-source software released under a lift Software License. Since development began in 2002, Dlib has grown to incorporate a good sort of tools. As of 2016, it contains software components for handling networking, threads, graphical user interfaces, data structures, algebra , machine learning, image processing, data processing , XML and text parsing, numerical optimization, Bayesian networks, and many other tasks. In recent years, much of the event has been focused on creating a broad set of statistical machine learning tools and in 2009 Dlib was published within the Journal of Machine Learning Research. Since then it has been used in a wide range of domains.

IV.RESULT AND SCREENSHOTS

Screenshots:

Home page
Blink Analysis
Micro Expression Analysis
Voice Energy Analysis
Voice Pitch Analysis
Voice Pitch Contour Analysis
Result Page

Result: First we have to upload the recorded video of question and answer session in the homepage and then, click on the analysis button at the right top corner of home page, we get all the analysis parameters and by clicking on the result button we get the result as shown in above screenshot.

V. CONCLUSIONS

These initial results indicate that it’s possible to correlate stress, lying (in the context of bluffing), and interest with a spread of physiological features. Using Artificial Neural Network, we’ve been ready to identify high stress situations to within about 82% accuracy. We can even discover lying with regarding seventy one accuracy. Essentially, we tend to demonstrate that we are ready to establish these events from easy aggregative physiological options non heritable throughout the amount of the events in question from non-invasively derived sensing.

VI REFERENCES

  1. [Caro 2003] M. Caro (2003) Caro‟s Book of Poker Tells, Car-doza Publishing
  2. [Lo & Repin 2002] A. Lo & D. Repin (2002) “The Psycho-physiology of Real-Time Financial Risk Processing”, Journal of Cognitive Neuroscience, 14:323-339
  3. [Lo et al. 2005] A. Lo, D. Repin, and B. Steenbarger (2005) “Fear and Greed in Financial Markets: A Clinical Study of Day-Traders”, American Economic Review
  4. [Madan et al. 2004] A. Madan, R. Caneel, and Pentland (2004) “GroupMedia: Distributed Multimodal Interfaces”, International Conference on MultiModal Interfaces
  5. [Marci 2002] C. Marci (2002) “Psychophysiologic Correlates of Empathy”, Abstract, Harvard Mysell Psychiatry Research Day, April, 2002
  6. [Picard 2001] R. Picard (2001) “Affective Medicine: Technology with Emotional Intelligence”, Chapter in Future of Health Technology, IOS Press
  7. [Picard et al. 2001] R. Picard, E. Vyzas, and J. Healey (2001) “Toward Machine Emotional Intelligence: Analysis of Affective Physiological State”, IEEE Transactions Pattern Analysis and Machine Intelligence, 23:10
  8. [Podlesny & Raskin 1977] J. Podlesny & D. Raskin (1977) “Physiological Measures and the Detection of Deception”, Psy-chol. Bull., 84(4):782-799
  9. [Steenbarger 2002] B. Steenbarger (2002) The Psychology of Trading: Tools and Techniques for Minding the Markets, John Wiley & Sons, Hoboken, NJ
  10. [Sung 2005] M. Sung (2005) “Non-Invasive Wearable Sensing Systems for Continuous Health Monitoring and LongTerm
  11. Barbour, A., (2002). The polygraph and the privacy tort.. Paper presented at the Annual Meeting of the Western States Communication Association (Long Beach, CA, March 2-5, 2002).
  12. Ekman, P., (1996). Why don’t we catch liars? Social Research, 63, 801-817.

Related posts

Android App for Managing Natural Calamity Affected Victims

admin

Netspam Detection using R

admin

Defence Mechnisms By Using Decoy Technology

admin

Leave a Comment