Download presentation
Presentation is loading. Please wait.
Published byDulcie Eaton Modified over 8 years ago
1
Detect Driver’s Emotion State By Using EEG Analysis Research Methodology Name: Ahmad Affandi Supli Matrix No: 812299 Prepared For: Dr. Farzana binti Kabir Ahmad
2
Background and Introduction The growing number of motor vehicles from time to time is always increasing. As a result, this has made some possibility the number of accidents or crashes are also increase. According to World Health Organization (WHO, 2013) in recent years around the world, the statistic have shown approximately about 3400 people die each day, Ten millions people are disable or injured in each year. Pedestrians, children, cyclists and old people are the most susceptible users on the road. There are so many attempts from the government and nongovernment in order to overcome this issue, such as by promoting good practices to use helmet, wear seat belt, and prevent from drinking and speeding through driving and so forth.WHO, 2013 For these reasons, many researchers had been trying to investigate this issue during the last decades. Lee (2008) stated that one of the prominent factors which are really pivotal in driving case is the behavior of drivers. Meanwhile, Mesken (2006) found in his study that emotion of driver is also has effect in affecting driver behavior. Another evident was showed from James (2000) work that stated when driver is angry; it can lead to disruption of performance while driving. Actually, many researchers, Lisetti and LeRouge (2004) and Zhiliang (2006) have been investigating for detecting emotion stated by tracing the alteration of human face expressions and also detecting from signals of physiological in real time. Eventually this research firstly proposed by (Qiang Ji et al., 2004) to continually make detection model for emotion states, which involved private information, information of background, and biosignals by using Bayesian Networks (BNs).
3
1.2 Problem Statement According to recent study that has been explained previously in introduction, the study proposed a detection model for driver emotion stated by using EEG with Bayesian Network (BNs), which involved the traffic situation and driver personality as an input prior factor. The results showed that the detection model was good in terms of performance. This study has stridden a step forward from the goal of the first study that developed using bio-signals for the purposed of the study However, this experiment to get those data samples were from particular videos driving simulation, which means the participant of experiments just watch the driving simulation without directly interacting in order to invoke their emotion (Fan et al., 2010). Therefore, this study focused on validating data samples which are coming from the real driver simulation in order to get data. In other words, the participant on experiments will be involved to drive throughout the graphical world of driver simulation. Moreover, the previous study also didn’t specify what kind of external environments were used in those videos. There are many external environment that could trigger when people driving such as advertisement, traffic jam or other possibility throughout driving in simulation world. This study will also focus on comparing the accuracy data which from any other methods beside Bayesian Network in order to see which is the best methods.
4
1.3 Significance of Study The significances of this study are as follows: Give insight of information in order to create in practice about developing driver assistant system which can contribute to decrease the number of crashes. The feedback from this study also will enrich another result in terms of type of data samples which are from the graphical world simulation to be validated from detection model of human emotion states. The results on this study will also give contribution for developing such us service robot which means as personal robot.
5
1.4 Research Questions The research questions are as follows: What kind of particular external environment type that will be used in driving simulation world? How to develop a prototype of simulation world that contains particular external environmental chosen? What are the other detection models that will be used in order to compare with Bayesin network model? What is the particular detection model that gives the highest accuracy to detect emotion states?
6
1.5 Research Objectives In this study, the aim of this study is to expand the experiment of detecting emotion’s driver states by using graphical simulation world. The specific objectives of this project are: To identify kind of particular external environment type that will be used in driving simulation world. To develop a prototype of simulation world that contains particular external environmental chosen. To identify the other detection models that will be used in order to compare with Bayesin network method. To determine which particular detection model that gives the highest accuracy to detect emotion states?
7
1.6 Problem Tree Human Emotions Facial Expressions, Prosody & Peripheral Neurons Lack of interpretation Unreliable EEG Not fit for people who want to conceal their feelings Lack of experiment to gain exact accuracy Less experiment Correlation is not consistent Not good enough
Similar presentations
© 2024 SlidePlayer.com. Inc.
All rights reserved.