• Healthcare Informatics Research
• /
• 제24권4호
• /
• pp.309-316
• /
• 2018

Objectives: Both the valence and arousal components of affect are important considerations when managing mental healthcare because they are associated with affective and physiological responses. Research on arousal and valence analysis, which uses images, texts, and physiological signals that employ deep learning, is actively underway; research investigating how to improve the recognition rate is needed. The goal of this research was to design a deep learning framework and model to classify arousal and valence, indicating positive and negative degrees of emotion as high or low. Methods: The proposed arousal and valence classification model to analyze the affective state was tested using data from 40 channels provided by a dataset for emotion analysis using electrocardiography (EEG), physiological, and video signals (the DEAP dataset). Experiments were based on 10 selected featured central and peripheral nervous system data points, using long short-term memory (LSTM) as a deep learning method. Results: The arousal and valence were classified and visualized on a two-dimensional coordinate plane. Profiles were designed depending on the number of hidden layers, nodes, and hyperparameters according to the error rate. The experimental results show an arousal and valence classification model accuracy of 74.65 and 78%, respectively. The proposed model performed better than previous other models. Conclusions: The proposed model appears to be effective in analyzing arousal and valence; specifically, it is expected that affective analysis using physiological signals based on LSTM will be possible without manual feature extraction. In a future study, the classification model will be adopted in mental healthcare management systems.

• Journal of Korea Society of Industrial Information Systems
• /
• 제26권1호
• /
• pp.1-10
• /
• 2021

This study proposed a Long-Short Term Memory network to consider changes in emotion over time, and applied an attention mechanism to give weights to the emotion states that appear at specific moments. We used 32 channel EEG data from DEAP database. A 2-level classification (Low and High) experiment and a 3-level classification experiment (Low, Middle, and High) were performed on Valence and Arousal emotion model. As a result, accuracy of the 2-level classification experiment was 90.1% for Valence and 88.1% for Arousal. The accuracy of 3-level classification was 83.5% for Valence and 82.5% for Arousal.

• Science of Emotion and Sensibility
• /
• 제17권3호
• /
• pp.57-64
• /
• 2014

In this paper, we verify the relation between elements (active and inactive) of Russell's emotional dimension ("A Circumplex Model") to propose a new representing method. Russell's emotional dimension expresses emotional words (happy, joy, sad, nervous, etc.) as a point on the two dimensions (Arousal and Valence). It is most commonly used in many filed such as Science of Emotion & Sensibility, Human-Computer Interaction (HCI), and Psychology etc. But other researchers have insisted that Russell's emotional dimension have to be modified because of its inherent problems. Such problems included the possibility of mixed feelings, the difference of emotion and sensibility, and the difference of Arousal axis and Valence axis. Therefore, we verify relationship of A Circumplex Model's elements (active and inactive) and find how to people express their Arousal feelings using survey. We finally propose new method to express emotion in Russell's emotional dimension. Using this method, we can solve Russell's problems and compensate other researches.

• Journal of Korean Library and Information Science Society
• /
• 제47권2호
• /
• pp.51-70
• /
• 2016

The purpose of this study is to analyze the semantic relation network and valence-arousal dimension through the words that describe emotions in Korean language. The results of this analysis are summarized as follows. Firstly, each emotion word was semantically linked in the network. This particular feature hinders differentiating various types of "emotion words" in accordance with similarity in meaning. Instead, central emotion words playing a central role in a network was identified. Secondly, many words are classified as two categories at the valence and arousal level: (1) negative of valence and high of arousal, (2) negative of valence and middle of arousal. This aspects of Korean emotional words would be useful to analyze emotions in various text data of books and document information.

• 한국HCI학회:학술대회논문집
• /
• 한국HCI학회 2009년도 학술대회
• /
• pp.707-710
• /
• 2009

In this paper, we introduce an interface for exploring music using emotional model. First, we survey arousal-valence factors of various music and calculate a correlation between audio fefatures of music and arousal-valence factors to build an AV model. Then, various music is aligned and arranged using the AV model and the user can explore music in this interface. To select the desired music more intuitively, we introduce new fade in/out function based on the location of the user's mouse point. We also offer several mode of selecting music so user can explore music using most suitable mode of interface. With our interface, the user can find the emotionally desired music more easily.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• 제35권2호
• /
• pp.64-70
• /
• 2012

This paper presents a statistical analysis method for the selection of electroencephalogram (EEG) electrode positions and spectral features to recognize emotion, where emotional valence and arousal are classified into three and two levels, respectively. Ten experiments for a subject were performed under three categorized IAPS (International Affective Picture System) pictures, i.e., high valence and high arousal, medium valence and low arousal, and low valence and high arousal. The electroencephalogram was recorded from 12 sites according to the international 10~20 system referenced to Cz. The statistical analysis approach using ANOVA with Tukey's HSD is employed to identify statistically significant EEG electrode positions and spectral features in the emotion recognition.

• The Journal of the Acoustical Society of Korea
• /
• 제42권1호
• /
• pp.75-82
• /
• 2023

The purpose of this study was to investigate listener's affective responses to different vocal registers and modes in terms of valence (i.e., negative to positive affect) and arousal (i.e., low to high energy level). The data were collected from four different conditions (i.e., higher and lower registers paired with major and minor modes). A total of 188 female college students participated in the survey online and rated their perceived valence and arousal levels on a visual analogue scale after listening to each excerpt. The two-way analysis of variance (ANOVA) was administered for data analysis. The results revealed that there were significant differences in the affective responses to the two vocal registers, showing that the arousal was more affected by the register than the valence. Secondly, mode had statistically significant impact on both valence and arousal while weighing more on valence. Further, there was significant interaction effect of vocal register and mode on valence, but not on arousal. Results also displayed that listeners had the most negative valence when listening to the excerpt of minor mode in higher register, while having the lowest arousal when listening to the excerpt of minor mode in lower register. These findings imply that it is important to consider the vocal range as well as the musical mode when selecting music for appreciation.

• Korean Journal of Cognitive Science
• /
• 제28권1호
• /
• pp.43-64
• /
• 2017

The purpose of the current study is to investigate how task-irrelevant affective priming affects higher cognitive function. In the study, we selected prime stimuli from International Affective Picture System(IAPS) and examined if they influence participants' performance of syllogistic reasoning task when they are task-irrelevant. In Experiment 1, arousal of IAPS stimuli was controlled while valence of the stimuli was manipulated. In Experiment 2, valence of IAPS stimuli was controlled while arousal of stimuli was manipulated. In both experiments, task-irrelevant affective primes were followed by syllogistic reasoning tasks consisting of three sentences and measured accuracies of task performance. The results showed that valence of affective prime affected logical validity of reasoning and belief-bias whereas arousal of affective primes did not yield any difference. That is, positive valence facilitated logical and analytic processing by reducing belief-bias while arousal did not affect reasoning task performance. These results suggest that dimensions of valence and arousal independently influence higher cognitive function.

• Korean Journal of Cognitive Science
• /
• 제20권2호
• /
• pp.183-195
• /
• 2009

We made Korean IAPS through measuring Korean university students' emotional response(arousal and emotional valence) about the whole 956 IAPS pictures made by Lang et al.(2005). In addition, we examined the emotional difference between American and Korean by comparing the response of original American IAPS and those of Korean IAPS. The results showed that both response of arousal and emotional valence in Korean were highly correlated with those in American respectively. In details, two groups showed differences as well as similarities. Korean showed higher arousal response than American, but in both groups women showed higher arousal response than men. When examining the emotional valence of positive, neutral, and negative stimuli categorized by American IAPS, Korean showed more modest emotional valence than American, and this group difference was the same in both men and women. In particular, Korean women showed more negative emotional valence than Korean men, but American women showed more extreme emotional valence than American men. These results suggest that there are some cultural and sex differences in the emotional response, and that researchers have to consider them when studying with IAPS stimuli.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 한국방송∙미디어공학회 2020년도 추계학술대회
• /
• pp.79-82
• /
• 2020

얼굴 영상으로부터 사람의 감정을 예측하는 연구는 최근 딥러닝의 발전과 함께 주목받고 있다. 본 연구에서 우리는 연속적인 변수를 사용하여 감정을 표현하는 dimensional model에 기반하여 얼굴 영상으로부터 감정 상태를 나타내는 지표인 valance/arousal(V/A)을 예측하는 딥러닝 네트워크를 제안한다. 그러나 V/A 예측 모델의 학습에 사용되는 기존의 데이터셋들은 데이터 불균형(data imbalance) 문제를 가진다. 이를 해소하기 위해, 우리는 오토인코더 구조를 가지는 얼굴 영상 생성 네트워크를 학습하고, 이로부터 얻어지는 균일한 분포의 데이터로부터 V/A 예측 네트워크를 학습한다. 실험을 통해 우리는 제안하는 얼굴 생성 오토인코더가 in-the-wild 환경의 데이터셋으로부터 임의의 valence, arousal에 대응하는 얼굴 영상을 성공적으로 생생함을 보인다. 그리고, 이를 통해 학습된 V/A 예측 네트워크가 기존의 under-sampling, over-sampling 방영들과 비교하여 더 높은 인식 성능을 달성함을 보인다. 마지막으로 기존의 방법들과 제안하는 V/A 예측 네트워크의 성능을 정량적으로 비교한다.