Abdul Wahab Abdul Rahman

Classification of EEG correlates on emotion using features from Gaussian mixtures of EEG spectrogram

This paper presents the classification of EEG correlates on emotion using features extracted by Gaussian mixtures of EEG spectrogram. This method is compared with three feature extraction methods based on fractal dimension of EEG signal including Higuchi, Minkowski Bouligand, and Fractional Brownian motion. The K nearest neighbor and Support Vector Machine are applied to classify extracted features. The 4 emotional states investigated in this paper are defined using the valence-arousal plane: two valence states (positive and negative) and two arousal states (calm, excited). The accuracy of system to classify 4 emotional states is investigated on EEG collected from 26 subjects (20 to 32 years old) while exposed to emotionally-related visual and audio stimuli. The results showed that the proposed feature extraction using Gaussian mixtures of EEG spectrogram yielded better classification results using the KNN classifier.

Systematic review of computational modeling of mood and emotion

In the recent years, more studies that aim to make computers understand, experience and respond to affects accordingly through computational models have been widely researched. Although many studies have defined and distinguished the words affect, mood and emotion, such terms are still used interchangeably. Thus, in this study, a systematic literature review was implemented to summarize and evaluate the current states of the arts on computational modeling of mood. From three online databases including IEEE Xplore, ScienceDirect and Springer Link, 825 scientific articles were extracted. Furthermore, through the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta Analyses) Statement, 9 articles were selected for the review. These resulting articles were reviewed based on several categories including the aim of the study, the study population, the measurement of mood, the basic emotion dimension, and proposed computational model, as well as evaluation. As a result, the systematic literature review has provided a good starting point in the study of the computational modeling of mood and emotion.

Characterizing autistic disorder based on Principle Component Analysis

Autism is often diagnosed during preschool or toddled age. This diagnosis often depends on behavioral test. It is known that individuals with autism have abnormal brain signals different from typical persons yet this difference in signals is slight that it is often difficult to distinguish from the normal. However, Electroencephalogram (EEG) signals have a lot of information which reflect the behavior of brain functions which therefore captures the marker for autism, help to early diagnose and speed the treatment. This work investigates and compares classification process for autism in open-eyed tasks and motor movement by using Principle Component Analysis (PCA) for feature extracted in Time-frequency domain to reduce data dimension. The results show that the proposed method gives accuracy in the range 90–100% for autism and normal children in motor task and around 90% to detect normal in open-eyed tasks though difficult to detect autism in this task.

Motor movement for autism spectrum disorder (ASD) detection

In this paper, we are looking at the differences between autistic and normal children in term of fine motor movement. Previous findings have shown that there are differences between autistic children and normal children when performing a simple motor movement tasks. Imitating a finger tapping and clinching a hand are two examples of a simple motor movement tasks. Our study had adopted one of the video stimuli for clinching the hand from Brainmarkers. 6 selected autistic children and 6 selected normal children were involved in this study. The data collection is using EEG device and will be analyzed using Gaussian mixture model (GMM) and Multilayer perceptron (MLP) as classifier to discriminate between autistic and normal children. Experimental result shows the potential of verifying between autistic and normal children with accuracy of 92%. The potential of using these techniques to identify autistic children can help early detection for the purpose of early intervention. Moreover, the spectrums of the signals also present big differences between the two groups.

Clustering Natural Language Morphemes from EEG Signals Using the Artificial Bee Colony Algorithm

We present a preliminary study on the use of a Brain Computer Interface (BCI) device to investigate the feasibility of recognizing patterns of natural language morphemes from EEG signals. This study aims at analyzing EEG signals for the purpose of clustering natural language morphemes using the Artificial Bee Colony (ABC) algorithm. Using as input the features extracted from EEG signals during morphological priming tasks, our experimental results indicate that applying the ABC algorithm on EEG datasets to cluster Malay morphemes produces promising results.

Speech emotion identification analysis based on different spectral feature extraction methods

Human speech communication will convey semantic information of the uttered word as well as the underlying emotion information of the interlocutor. Emotion identification is important, as it could enhance many applications added-features that can improve human computer interaction aspect. Such improvement surely can help to retain customer satisfaction and loyalty in the long run and serves as an attraction factor for a new customer. Although many researchers have used many approaches to recognize emotion from speech, no one can claim superiority of their findings. This is because different feature extraction methods coupled with various classifiers may produce different performance depending on the data used. This paper presents a comparative analysis of the speech emotion identification system using two different feature extraction methods of Mel Frequency Cepstral Coefficient (MFCC) and Linear Prediction Coefficient (LPC) coupled with Multilayer Perceptron (MLP) classifier. For further exploration, different numbers of MFCC filters are employed to observe the performance of the proposed system. The results indicate that MFCC-40 gives slightly better performance compared to the other MFCC coefficients in the Berlin EMO-DB and NTU_American whereas the MFCC-20 performs well for NTU_Asian. It is also observed that MFCC consistently performed better than LPC in all experiments, which are in-line with many reported findings. Such understanding can be extended to further study speech emotion in order to develop more robust and least error system in the future.

Stress Assessment While Listening to Quran Recitation

Stress and anxiety are one of the most widespread problems presently, stress treatment has been featured in many researches. The use of Quran offers a substantial help in treating stress. The purpose of this study is to examine the various aspects and perspectives of human emotions while listening to Quran Recitation. This study aims to identify and select verses that have more psychological impact than other verses and to identify the most Quran reciter that respondents believe his voice brings calmness and tranquility to their mind. Quantitative and qualitative methods were adopted. An online distributed questionnaire was sent to the academic staffs of all the Islamic faculties in the Malaysian public universities, five subjects participated in EEG experiment to identify their emotions while listening to these Quran verses. Five Quran verses and one reciter were identified, while experiment indicates that the subjects are more relieved and relaxed when listening to these Quran verses.

Electrocardiogram Identification: Use a Simple Set of Features in QRS Complex to Identify Individuals

This paper presents a Multilayer Perception Neural Network developed to identify human subjects using electrocardiogram (ECG) signals. We use the amplitude values of Q, R and S as a features for our experiments. In this study, a total of 87 dataset were collected among 14 subjects from the Physikalisch-Technische Bundesanstalt (PTB) database. Out of the 14 subjects, Q-R-S feature points were taken from different day and time sessions to perform classification with MLP. Out of this data, 66 % is used as training dataset while the remaining 34 % is used for testing. Our method yields 96 % accuracy and demonstrates that the use of three fiducial points is sufficient to identify a subject despite the common practice of taking more feature points.

EEG-based Emotion Recognition while Listening to Quran Recitation Compared with Relaxing Music Using Valence-Arousal Model

Relaxation and calmness are two emotions that people always seek for. One popular method people used to do in order to reduce their level of tension and pressure is listening to some types of relaxing music. On the other hand, Quran is Allahs words that are ultimately given to us human to benefit of. Although, Muslims are strongly believed that listening to Quran or reading it brings them to comfort, pleasure and confidence. Scientific evidence is still required to prove that scientifically. Human emotion can be recognized from voice, text, facial expression or body language. But those methods are susceptible to change and are not really accurate. Recently, electroencephalograms (EEG) allowed researchers to evoke the inner emotions. This paper aims to study human emotions while listening to Quran recitation compared with listening to relaxing music. To evoke emotions, some stimuli should be used, in this research we implemented International Affective Picture System (IAPS) database. And for the emotion classification technique we followed two-dimensional Arousal-Valence emotion model. Finally the emotion model was implemented to recognize four basic emotions Happy, Fear, Sad and Calm with an average accuracy of 76.81 %. The data collected while listening to Quran and music were tested and the result generally showed that both Quran and Music are classified more into positive valence.

Optimizing human memory: An insight from the study of Al Huffaz

We investigated the memory control processes used by a group of people who commit to memory the entire word text of the Quran with complete accuracy. The findings from interviews conducted, revealed the importance of several control processes such as rehearsal, motivation or interest and self-discipline. More importantly, it has been found that it is maintenance and not elaborative rehearsal that plays the most important role in the memorization of the Quran. These findings will help to form a better understanding of the cognitive basis underlying human memory. This knowledge can be incorporated with technology for the purpose of optimizing human potential. More importantly, it gives implications to learning process at large and opens new opportunities and ways to serve the Ummah better.