Introduction: In this study, we propose a new adaptive method for fusing multiple emotional modalities to improve the performance of an emotion recognition system.Method: Three-channel forehead biosignals, along with peripheral physiological measurements (blood volume pressure, skin conductance, and interbeat intervals), were utilized as emotional modalities. Six basic emotions, i.e., anger, sadness, fear, disgust, happiness, and surprise were elicited by displaying preselected video clips for each of the 25 participants in the experiment. In the proposed emotion recognition system, recorded signals with the formation of three classification units identified the emotions independently. The results were then fused using the adaptive weighted linear model to produce the final result. Each classification unit is assigned a weight that minimizes the squared error of the ensemble system.Results: The results showed that, the proposed fusion method outperformed all individual classifiers and emotion systems that were designed based on feature level fusion and classifiers fusion using the majority voting method. Using the support vector machine (SVM) classifier, an overall recognition accuracy of 88% was obtained in identifying the intended emotional states. Also, applying only the forehead or the physiological signals in the proposed fusion scheme indicates that designing a reliable emotion recognition system is feasible without the need for additional emotional modalities.Conclusion: The results suggest using adaptive fusion of classification units in the design of multimodal emotions recognition system.

Introduction: Emotions play an important role in health, communication, and interaction between humans. The ability to recognize the emotional status of people is an important indicator of health and natural relationships. In DEAP database, electroencephalogram (EEG) signals as well as environmental physiological signals related to 32 volunteers are registered. The participants in each video were rated in terms of level of arousal, capacity, liking/disliking, proficiency, and familiarity with the video they watched. Method: In this study, a practical empirical method was adopted to classify capacity, arousal, proficiency, and interest by ranking the features extracted from signals using algorithms on EEG signals and environmental physiological signals (such as electromyography (EMG), electrooculography (EOG), galvanic skin response (GSR), respiration rate, photoplethysmography (PPG), and skin temperature. After initializing the signals from the database and pre-processing them, various features in the time and frequency domain were extracted from all signals. In this study, SVM and KNN classifiers, K-means clustering algorithm, and neural networks, such as PNN and GRNN were used to identify and classify emotions. Results: It was indicated in this study that the results of the classification of emotions using various methods and classifiers were well-established with high accuracy. The best accuracy results were obtained by applying the proposed method using SVM classifier based on features extracted from environmental signals (85. 5%) and EEG signals (82. 4%). Conclusion: According to the results of the classification of emotions in this study, the proposed algorithm provides relatively better results compared with previous similar methods.

In this study, we propose decision level fusion of multimodal physiological signals to design an affect identification system using the MIT database. Four types of physiological signals, including blood volume pressure (BVP), respiration rate (RSP), skin conductance and facial muscles activities (fEMG) were utilized as affective modalities. To collect the above-mentioned database, researchers used personalized imagery to elicit the desired affective states from a single subject and recorded the corresponding physiological signals simultaneously. In this study, the best subset of features for each signal was determined using previously calculated time and frequency domain features. To this end, sequential floating forward selection (SFFS) and RELIEF feature selection algorithms were evaluated. A new feature set, formed by concatenating the selected features, was partitioned into three subsets. Each subset was then fed into a classifier to identify the desired affective states. The majority voting method was applied to fuse the results obtained by the subsystems. Three types of classification methods, namely SVM, LDA and KNN were evaluated to design an affect identification system. The results showed remarkable performance from the system in identifying the desired scenarios with an acceptable accuracy and speed of response. Using the RELIEF feature selection method, along with SVM as a classifier, an overall recognition accuracy of 93.8% was obtained, which is better than the results reported with the use of the above-mentioned database so far.

This research project presents a comprehensive methodology for stress identification by combining subjective self-report data and objective physiological signals. The proposed system employs a carefully designed questionnaire, tailored to different age groups, to enhance accuracy in stress assessment. Subjects respond to the questionnaire, providing valuable insights into their emotional well-being. Subsequently, physiological data is collected using an infrared (IR) sensor positioned beneath the wrist, close to the artery. The pulse data obtained is meticulously converted into a CSV file, allowing for efficient preprocessing. The preprocessing phase ensures the integrity of the data, preparing it for machine learning (ML) analysis. The study harnesses ML techniques, specifically SVM (Support Vector Machines) & KNN (K-Nearest Neighbors), to classify stress levels based on the pre-processed data. Through feature extraction, relevant patterns are identified, contributing to the accurate characterization of stress states. This integrative approach offers a robust framework for stress assessment, taking into account both subjective and physiological dimensions. Results demonstrate promising accuracy levels: Support Vector Machine (SVM) Reached a level of precision of 0.98 (+/- 0.20), Decision Tree showed 0.93 (+/- 0.30), and K-Nearest Neighbors (KNN) reached 0.88 (+/- 0.44). It also implements the voting classifier for improved performance of 98.6% of accuracy. These findings underscore the effectiveness of the proposed methodology in accurately identifying stress levels. Integrating subjective insights with objective physiological data not only enhances stress identification but also offers a comprehension of the intricate correlation between mental states and physiological reactions. This comprehensive strategy holds substantial implications across diverse domains such as healthcare, psychology, and human-computer interaction.

Stress has affected human’ s lives in many areas, today. Stress can adversely affect human’ s health to such a degree as to either cause death or indicate a major contributor to death. Therefore, in recent years, some researchers have focused to developing systems to detect stress and then presenting viable solutions to manage this issue. Generally, stress can be identified through three different methods including (1) Psychological Evaluation, (2) Behavioral Responses and finally (3) physiological signals. physiological signals are internal signs of functioning the body, and therefore nowadays are commonly used in various medical and non-medical applications. Since these signals are correlated with the stress, they have been commonly used in detection of the stress in humans. Photoplethysmography (PPG) and Galvanic Skin Response (GSR) are two of the most common signals which have been widely used in many stress related studies. PPG is a noninvasive method to measure the blood volume changes in blood vessels and GSR refers to changes in sweat gland activity that are reflective of the intensity of human emotional state. Design and fabrication of a real-time handheld system in order to detect and display the stress level is the main aim of this paper. The fabricated stress monitoring device is completely compatible with both wired and wireless sensor devices. The GSR and PPG signals are used in the developed system. The mentioned signals are acquired using appropriate sensors and are displayed to the user after initial signal processing operation. The main processor of the developed system is ARM-cortex A8 and its graphical user interface (GUI) is based on C++ programming language. Artificial Neural Networks such as MLP and Adaptive Neuro-Fuzzy Inference System (ANFIS) are utilized to modeling and estimation of the stress index. The results show that ANFIS model have a good accuracy with a coefficient of determination values of 0. 9291 and average relative error of 0. 007.

Detecting and controlling stress levels in drivers is especially important to reduce the potential risks while driving. Accordingly, in this study, a detection system was presented to identify four levels of stress (low, neutral, high and very high) in drivers based on physiological signals. The proposed method used the drivedb database, which includes the recording of physiological signals from 17 healthy volunteers while driving on specific routes on city streets and highways. A set of statistical and entropy features along with morphological features that were calculated only for the ECG signals, were used. The calculated features were applied as inputs to the classification units to detect stress levels. Support vector machine (SVM), k nearest neighbors (kNN) and decision tree (DT) were evaluated as classification methods. The main purpose of this study was to improve the accuracy of stress level detection using the idea of classifiers fusion. To achieve this goal, the combination of individual classification units, each of which used only the features of one of the ECG, EMG and GSR signals, was performed by the Demster-Shafer method. Using genetic algorithm as feature selection method, SVM classifier and Dempster-Shafer fusion strategy, the best stress detection accuracy of 96. 9% was obtained. While the highest detection accuracy among individual classifiers was 75% and obtained by a subsystem that used ECG features. The results show significant performance of the proposed method compared to previous studies that used the same dataset.

Human stress is a physiological tension that appears when a person responds to mental, emotional, or physical chal-lenges. Detecting human stress and developing methods to manage it, has become an important issue nowadays. Auto-matic stress detection through physiological signals may be a useful method to solve this problem. In most of the earlier studies, long-term time window was considered for stress detection. Continuous and a real-time representation of the stress level are usually done through one physiological signal. In this paper, a real-time stress monitoring system is pro-posed which shows the user a new signal for feedback stress level. This signal is the combination of weighted features of galvanic skin response and photoplethysmography signals. The features are defined in 20-sec time windows. Correlation feature selection and linear regression methods are used for feature selection and feature combination, respectively. Furthermore, a set of experiments was conducted to train and test of the proposed model. The proposed model can represent the relative stress level perfectly and has 79% accuracy for classifying the stress and relaxation phases into two categories by a determined threshold.