EEG-based emotion recognition using machine learning techniques

Lan, Zirui

doi:10.32657/10220/46340

Cited by 2 publications

(2 citation statements)

References 160 publications

(258 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Moreover, the stimuli in the IADS database have been excerpted from real-life events or scenarios. In order to minimize the possible variance of response from participants from different socio-cultural backgrounds, these scenarios were carefully selected while developing the IADS database [60]. For example, in order to induce positive pleasant emotions, the sounds of bird's merry chirping, stream water flowing, or children's laughter are used.…”

Section: Paradigm3-si (Experiments Performed While Listening Sounds)mentioning

confidence: 99%

Comparing Neural Correlates of Human Emotions across Multiple Stimulus Presentation Paradigms

Masood

Farooq

2021

Brain Sciences

View full text Add to dashboard Cite

Most electroencephalography (EEG)-based emotion recognition systems rely on a single stimulus to evoke emotions. These systems make use of videos, sounds, and images as stimuli. Few studies have been found for self-induced emotions. The question “if different stimulus presentation paradigms for same emotion, produce any subject and stimulus independent neural correlates” remains unanswered. Furthermore, we found that there are publicly available datasets that are used in a large number of studies targeting EEG-based human emotional state recognition. Since one of the major concerns and contributions of this work is towards classifying emotions while subjects experience different stimulus-presentation paradigms, we need to perform new experiments. This paper presents a novel experimental study that recorded EEG data for three different human emotional states evoked with four different stimuli presentation paradigms. Fear, neutral, and joy have been considered as three emotional states. In this work, features were extracted with common spatial pattern (CSP) from recorded EEG data and classified through linear discriminant analysis (LDA). The considered emotion-evoking paradigms included emotional imagery, pictures, sounds, and audio–video movie clips. Experiments were conducted with twenty-five participants. Classification performance in different paradigms was evaluated, considering different spectral bands. With a few exceptions, all paradigms showed the best emotion recognition for higher frequency spectral ranges. Interestingly, joy emotions were classified more strongly as compared to fear. The average neural patterns for fear vs. joy emotional states are presented with topographical maps based on spatial filters obtained with CSP for averaged band power changes for all four paradigms. With respect to the spectral bands, beta and alpha oscillation responses produced the highest number of significant results for the paradigms under consideration. With respect to brain region, the frontal lobe produced the most significant results irrespective of paradigms and spectral bands. The temporal site also played an effective role in generating statistically significant findings. To the best of our knowledge, no study has been conducted for EEG emotion recognition while considering four different stimuli paradigms. This work provides a good contribution towards designing EEG-based system for human emotion recognition that could work effectively in different real-time scenarios.

show abstract

Section: Paradigm3-si (Experiments Performed While Listening Sounds)mentioning

confidence: 99%

Comparing Neural Correlates of Human Emotions across Multiple Stimulus Presentation Paradigms

Masood

Farooq

2021

Brain Sciences

View full text Add to dashboard Cite

show abstract

“…Each proposed a framework to identify the major components involved in the learning system. First, Lan [12] proposed a framework for machine learning algorithms with four basic components of a personalized learning system (PLS) that involved learning analytics, content analytics, grading and feedback, and scheduling. This research integrated learning resources for math composed of textbooks, lecture notes, and homework assignments as data input into a PLS.…”

Section: Related Workmentioning

confidence: 99%

Exploring the Potential of Deep Learning for Personalized Learning Environments

Junus,

Brophy

2023 ASEE Annual Conference &Amp; Exposition Proceedings

View full text Add to dashboard Cite

This study investigates the current potential for artificial intelligence (AI) to support personalized learning (PL). Personalized learning can provide a customized learning environment to support student learning processes based on individual needs, competencies, and interests. One way to conduct personalized learning is by using a recommender system that employs deep learning, an AI technique. To date, a limited number of researchers have discussed the application of deep learning methods to develop advanced recommenders in personalized learning environments. This study examines the literature that describes deep learning as a recommender system to support personalized learning environments. This initial phase of the project seeks to synthesize the issues and opportunities associated with personalized learning experiences and the potential of using deep learning to support the process. Because the topic intersects the education and information technology (IT) fields, we selected three databases for this literature review project: Scopus, ERIC, and Engineering Village. We used the phrase "deep learning recommender system for personalized learning environments" as our search string. We focused only on papers that experts had evaluated in the field to ensure accuracy. Therefore, terms such as "peer review," "literature review," and "systematic review" were added to the original search string. The initial search results included 409 documents. After applying inclusion/exclusion criteria, 20 papers emerged as the focus of this study. Thematic analysis was used to look for various themes to identify how deep learning methods are used in education and their potential to inform personalized learning environments. The analysis process utilized Mendeley and NVivo to quickly capture themes by focusing on six features to peruse within the articles. The features involved research questions, goals of studies, research methodology, research design, primary outcomes, and limitations. We then generated three themes from the six features in the analysis phase of the 20 papers. The first theme categorized the type of study into primary and secondary studies. These categories identify the types of studies that employed deep learning methods in the development of a recommender system and their integration with personalized learning. The second theme, recommender system (RS) techniques, highlighted the AI methodologies most frequently utilized in previous research. And the third theme was a list of e-learning platforms that applied RS for personalized learning. The main findings revealed that the deep learning method was effective in big data analysis due to its ability to forecast students' achievements, behaviors, and future paths. Therefore, we considered that deep learning could be widely applied as a technique to develop recommender systems to support personalized learning environments. Furthermore, because we found that only a few studies have investigated the implementation of this AI technology, researchers will have a great...

show abstract

EEG-based emotion recognition using machine learning techniques

Cited by 2 publications

References 160 publications

Comparing Neural Correlates of Human Emotions across Multiple Stimulus Presentation Paradigms

Comparing Neural Correlates of Human Emotions across Multiple Stimulus Presentation Paradigms

Exploring the Potential of Deep Learning for Personalized Learning Environments

Contact Info

Product

Resources

About