Deep Learning-Based Emotion Recognition from Real-Time Videos

Zhou, Wei; Cheng, Justin; Lei, Xingyu; Beneš, Bedřich; Adamo, Nicoletta

doi:10.1007/978-3-030-49062-1_22

Cited by 10 publications

(5 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Emotion extraction from videos has also been investigated by applying deep learning. In [86], CNNs were applied to extract emotions from videos and audio streams simultaneously. Emotions have also been extracted from videos by applying a convolutional deep belief network, which achieved better recognition accuracies than the SVM baselines in multimodal scenarios [74].…”

Section: Video Data and Emotion Classificationmentioning

confidence: 99%

Emotion Classification from Speech and Text in Videos Using a Multimodal Approach

Caschera

Grifoni

Ferri

2022

MTI

View full text Add to dashboard Cite

Emotion classification is a research area in which there has been very intensive literature production concerning natural language processing, multimedia data, semantic knowledge discovery, social network mining, and text and multimedia data mining. This paper addresses the issue of emotion classification and proposes a method for classifying the emotions expressed in multimodal data extracted from videos. The proposed method models multimodal data as a sequence of features extracted from facial expressions, speech, gestures, and text, using a linguistic approach. Each sequence of multimodal data is correctly associated with the emotion by a method that models each emotion using a hidden Markov model. The trained model is evaluated on samples of multimodal sentences associated with seven basic emotions. The experimental results demonstrate a good classification rate for emotions.

show abstract

Section: Video Data and Emotion Classificationmentioning

confidence: 99%

Emotion Classification from Speech and Text in Videos Using a Multimodal Approach

Caschera

Grifoni

Ferri

2022

MTI

View full text Add to dashboard Cite

show abstract

“…The majority of FER approaches investigate human emotions under seven primary emotional states (Zhou et al, 2020), which are anger, disgust, fear, sadness, happiness, surprise, and neutral. In the scope of this study, the main focus is the assessment of students' engagement and impressions about the content covered during an online lecture.…”

Section: The Circumplex Model Of Affectmentioning

confidence: 99%

“…In this study, we present EduFERA, which is a real-time student emotion assessment approach that aims to improve the online lecturing experience. Inspired by the effectiveness of deep learning in correlated computer vision tasks (Farrell et al, 2019;Zhou et al, 2020;Zeng et al, 2020), the proposed system utilizes deep neural network architectures for face detection and FER. Several experiments are carried out on a common, realworld dataset.…”

Section: Introductionmentioning

confidence: 99%

EduFERA: A Real-Time Student Facial Emotion Recognition Approach

MOUHEB¹,

Yürekli²,

DERVİSBEGOVİC³

et al. 2022

European Journal of Science and Technology

View full text Add to dashboard Cite

The use of video conferencing tools in education has increased dramatically in recent years. Especially after the COVID-19 outbreak, many classes have been moved to online platforms due to social distancing precautions. While this trend eliminates physical dependencies in education and provides a continuous educational environment, it also creates some problems in the long term. Primarily, many instructors and students have reported issues concerning the lack of emotional interaction between participants. During in-place education, the speaker receives immediate emotional feedback through the expressions of the audience. However, it is not possible to fully utilize this valuable feedback in online lectures since current tools can only display a limited number of faces on the screen at a time. In order to alleviate this problem and promote the online education experience one step closer to in-place education, this study presents EduFERA that provides a real-time emotional assessment of students based on their facial expressions during video conferencing. Empirically, several state-of-the-art techniques have been employed for face recognition and facial emotion assessment. The resulting optimal model has been deployed as a Flask Web API with a user-friendly ReactJS frontend, which can be integrated as an extension to current online lecturing systems.

show abstract

“…This section will provide many significant works in recognizing faces containing three main parts: face recognition, discovering the features, and facial passivity [10]. Nowadays, many researches are done based on the deep learning technique to find the passivity and the emotions during distance education systems by analyzing the texts, and analyzing the face [11]. Table 1 summarizes many works related to this topic.…”

Section: Literature Reviewmentioning

confidence: 99%

Artificial Intelligence Techniques for Distance Education: A Systematic Literature Review

Aljarrah¹,

Ababneh²,

Karagozlu³

et al. 2021

TEM Journal

View full text Add to dashboard Cite

In the current era, education, like other fields, relies heavily on big data. Moreover, artificial intelligence, including affective computing, is one of the most essential and popular technologies adopted by educational institutions to process and analyze big data. In this systematic review, many previous research types related to improving educational systems using artificial intelligence techniques were studied, such as: deep learning, machine learning, and affective computing. This systematic review aims to identify the gaps in students' emotional understanding in distance education systems. The world has recently witnessed the spread of educational processes for distance learning, especially in the university and the enormous open online courses (MOOCs). Besides, the COVID-19 pandemic has been involved in changing all educational processes to a distance learning system. The results indicated that these systems recorded a high success rate. However, the teacher does not fully understand the student’s emotional state during the educational session. It also lacks monitoring or monitoring during the electronic exams, which are electronic exams. So, it is a widespread problem in distance learning.

show abstract

Deep Learning-Based Emotion Recognition from Real-Time Videos

Cited by 10 publications

References 46 publications

Emotion Classification from Speech and Text in Videos Using a Multimodal Approach

Emotion Classification from Speech and Text in Videos Using a Multimodal Approach

EduFERA: A Real-Time Student Facial Emotion Recognition Approach

Artificial Intelligence Techniques for Distance Education: A Systematic Literature Review

Contact Info

Product

Resources

About