AI-driven emotion recognition for adaptive learning in UAE educational institutions

Abstract

An AI-driven emotion recognition system for learning adaptation in educational institutions, comprehensive: • an emotion recognition unit that tracks students' facial expressions in real time and categorizes the resulting expressions into a specific emotional class from a given set using a Convolutional Neural Network (CNN) trained on a large emotion dataset; • a real-time processing unit that extracts facial features and dynamic emotional characteristics of students through adaptive filter algorithms; • an LMS integration unit, adapted to combine discovered emotional knowledge with online and offline learning environments for customizable learning adjustments; • a personalization engine programmed to intelligently adapt training content, response approaches, and training materials according to the identified emotional states, so that each learner receives individually tailored and more intensive learning experiences; • a feedback loop that sends educators real-time alerts and recommended videos to address emotional feedback or lesson gaps, allowing teachers to tackle negative signals in the moment (e.g., frustrated face, etc.); • AI data storage and adaptive learning model unit that stores anonymized data for emotion recognition, updates the AI ​​model and improves it through learning over time, and ensures that the use of the information is in compliance with data protection regulations; • A user interface module specifically designed for educators, providing a visual overview of students' emotional patterns and real-time analysis of their emotions and engagement as a basis for important decisions that lead to improvements in teaching processes.

Description

Field of invention

[0001] The present invention relates to applications based on artificial intelligence (AI) and, in particular, improves adaptive learning in educational institutions in the UAE through AI-based emotion recognition. Specifically, it proposes an AI-based system based on deep learning approaches, especially convolutional neural networks (CNNs), to recognize students' emotions in real time and adapt teaching methods accordingly to increase student understanding and engagement. Background of the invention:

[0002] This shift is largely due to the rapid development of artificial intelligence (AI) and machine learning (ML) technologies, which are gaining traction in various fields such as healthcare, finance, and education. Adaptive learning is one of the most exciting AI applications in education. It allows instructional content to be tailored to the needs, learning pace, and cognitive intelligence of individual students. However, traditional adaptive learning systems rely primarily on academic performance indicators such as test scores, participation rates, and learning progress, neglecting a crucial factor influencing learning: students' emotions.

[0003] Emotions are an essential part of the learning process. Cognitive psychology and neuroscience research have shown that emotional states influence attention, memory, and problem-solving abilities. Positive emotions studied include joy, enthusiasm, and excitement. Positive emotions help students better understand learning content, engage with it more effectively, and be more motivated. Negative emotions such as frustration, anxiety, and boredom, on the other hand, can impair cognitive performance, shorten attention spans, and lead to a lack of motivation. Immediately understanding and addressing students' emotions can improve learning outcomes, facilitate knowledge recall, and optimize lesson planning.

[0004] In response, the following invention proposes an artificial intelligence emotion recognition system aimed at educational institutions in the UAE. Based on advancements in state-of-the-art facial expression recognition (FER) technologies, this system combines adaptive learning structures to provide instant insights into students' emotional states. The developed system classifies seven basic emotions—anger, disgust, fear, joy, sadness, surprise, and neutrality—using deep learning models trained on FER2013 and other extensive datasets. By leveraging AI to capture students' emotional states, teachers can adapt their teaching methods in real time, creating dynamic, engaging, and personalized learning experiences.

[0005] Even as technology permeates the classroom, traditional teaching is still based on standardized methods. Some challenges of conventional learning environments include a uniform design that fails to consider the cognitive and emotional states of diverse students, the lack of real-time emotional feedback, limited technology integration, the diverse demographic makeup of the student body, and the importance of learning through emotional states. Because many educational institutions lack flexible curricula, they often struggle to motivate students, regardless of the number of group or active learning plans they employ. Furthermore, because teachers have limited ability to adapt their teaching strategies, they often lack the tools to capture students' emotional states in real time.Most edtech platforms focus on delivering and evaluating content (not emotions). In multicultural environments like UAE schools, students often come from diverse linguistic and cultural backgrounds, requiring a tailored and holistic approach to their education. Negative emotions can slow students' information processing, while positive emotions can increase engagement and motivation to learn.

[0006] This development is a unique, AI-based facial expression recognition system built on adaptive learning techniques. It performs emotion recognition using convolutional neural networks (CNNs) trained on extensive datasets, including FER2013. The CNN architecture incorporates multiple convolutional layers, batch normalization, dropouts, and various activation functions to effectively classify emotions. Key processes include data preprocessing, feature extraction, and real-time emotion classification. In face-to-face classrooms, the system processes facial features and microexpressions to accurately classify emotions. In virtual learning environments, it analyzes students' facial expressions in real time using camera streams or webcam feeds.

[0007] After recognizing students' emotional states, the learning management system (LMS) and digital learning system are integrated, enabling dynamic adaptation of teaching methods. This includes tailoring content to recognized emotional states, real-time notifications for teachers, and automated intervention mechanisms. Potential anxiety and disinterest among students present a challenge that the system can address. For example, if signs of frustration or disinterest are detected, motivational messages, interactive activities, alternative explanations, etc., are stored.

[0008] Through the use of machine learning techniques, the system can continuously improve its performance. Furthermore, it allows the model to learn continuously and enhance its predictive accuracy and adaptability by incorporating newly collected data from classrooms in the UAE. Contextualization and incremental model training are part of the learning pipeline. This process makes the model more effective as it learns from new data. Even better, the cultural and contextual factors learned by the students in the UAE are integrated into the emotion recognition process for this specific group, thus considering all factors relevant to emotion analysis. Argumentative essay.

[0009] This innovative invention offers numerous benefits that can contribute to improved education in UAE educational institutions. These include personalized learning experiences, increased student engagement, immediate support for teachers, seamless integration with existing platforms, cultural relevance, and enhanced learning outcomes. The system recognizes students' emotional states and adapts teaching strategies to their individual needs, thereby maximizing motivation and focus. AI-driven interventions keep students engaged and provide teachers with actionable insights that enable dynamic lesson planning adjustments. The system is applicable to both online and in-person instruction, making it a versatile tool for the 21st century. To improve accuracy and inclusivity, the model is trained using a region-based dataset.Addressing the emotional needs of students leads to learning success and academic achievement.

[0010] The use of AI-powered emotion recognition in adaptive learning represents a significant advancement in modern education. This invention combines the best of both worlds by merging deep learning techniques with real-time emotion recognition, providing a powerful solution for educational institutions in the UAE. This real-time adaptability of the teaching process to students' emotions contributes to a more inclusive, engaging, and ultimately more efficient learning experience. This comes as the UAE continues to make progress toward achieving its goal of AI-based innovation across education, aligning with its national strategic agenda and inspiring the next generation of learners to embrace cutting-edge technology. Summary of the invention:

[0011] Embodiments of the present invention relate to an AI-based emotion recognition system that enables adaptive learning in schools and universities, in particular an AI-based emotion recognition system in the UAE. It utilizes deep learning-based facial expression recognition (FER) technology to determine the emotional state of students in real time and to adjust the teaching style accordingly in order to make lessons more effective.

[0012] This AI-based solution utilizes convolutional neural networks (CNNs) and is trained on extensive emotion recognition datasets such as FER2013 to classify seven basic emotions: anger, disgust, fear, joy, sadness, surprise, and neutrality. By combining computer vision and deep learning, the system is trained to capture and interpret live facial information from students in both physical and virtual classrooms. This provides teachers with immediate feedback on their students' emotional responses, allowing them to adjust their lesson plans accordingly to ensure an optimal learning experience.

[0013] The systems integrate seamlessly with existing LMS and intelligent learning platforms, providing automated feedback and student-tailored content recommendations. They utilize sophisticated data preprocessing methods, such as data normalization and augmentation, to promote the accurate and efficient classification of emotions across diverse learning environments. Furthermore, the model facilitates this cycle of performance improvement because classrooms in the UAE provide the system with current and relevant data.

[0014] The AI-based emotion recognition system enables real-time interventions, creating a personalized and inclusive learning environment. It supports synchronous and asynchronous learning models and allows for real-time monitoring of emotions in both live and recorded lessons. Furthermore, the system can track student mood trends over time, enabling teachers and school administrators to use the data to drive improvements in curriculum design and teaching strategies.

[0015] This invention enables real-time, emotion-based interventions, creating an inclusive, student-centered learning experience that improves overall engagement and academic performance. The system aligns with the UAE's national strategies for using AI in meetings and offers a data-driven, scalable approach for modern classrooms. Its impact extends beyond traditional teaching, enabling hybrid and fully digital learning paradigms and making it a transformative tool for the future of education. Furthermore, this system can be more broadly deployed in corporate training, mental health monitoring, and professional expertise, representing a groundbreaking development in AI-powered education. Description of the invention

[0016] These and other advantages, features and aspects of the present invention will be better understood with reference to the detailed description below of the exemplary embodiments in conjunction with the accompanying figures, in which the same reference numerals refer to similar parts in the different views.

[0017] AI-powered emotion recognition system for adaptive learning in UAE educational institutions, as described in Fig. depicted.

[0018] The block diagram representation of the AI-driven emotion recognition system for adaptive learning is in Fig. The system 100 comprises: • An emotion recognition unit 102, consisting of a facial expression recognition module 102(a), supported by a convolutional neural network (CNN), trained on a huge dataset (e.g., FER2013). This module can identify students' emotions and categorize them into seven basic emotions such as anger, disgust, fear, happiness, sadness, surprise, and neutrality. The facial expression recognition module processes the live stream and monitors facial features and microexpressions of students in the classroom; • A real-time processing unit 104 that processes the captured emotional data and provides real-time insights for educators. The following unit utilizes various deep learning inference techniques that dynamically analyze student behavior patterns. Previously, the data was analyzed manually, but in this new system, the real-time processing unit 104(a) includes a feature extraction unit and an adaptive filtering algorithm that accepts inputs based on the collected raw data, resulting in more accurate results; • A learning management system integration unit 106, specifically designed to synchronize this emotion recognition data with both online and offline educational platforms. This unit collects and transmits emotional indicators in real time to teachers via a user interface, thus enabling adaptive teaching strategies. It is easily compatible with current smart education systems and allows for automatic, emotion-based lesson adjustments. [The text then abruptly shifts to a different topic:] of students captured on camera; • A personalization engine 108 that adapts the delivery of learning content to the feelings of the students. This engine uses recommendation algorithms based on artificial intelligence to personalize learning content and adjust the difficulty level, pace, or presentation style to the individual emotional state of the learner. The personalization engine 108(a) also uses historical emotion data to guide the user to the appropriate interactive activity; • A feedback mechanism 110 provides real-time recommendations to educators and student activities. Alerts and lesson adjustments are triggered by signs of emotions such as boredom or frustration; the feedback mechanism unit 110(a) enables a timely response. It also facilitates long-term trend analysis, allowing teachers to adapt their educational strategies in response to recurring emotional patterns. • Unit 112 Data Storage and Adaptive Learning Model, or simply Unit 112 for securely storing student emotion data and optimizing your AI model. It keeps the models improving, anonymizes your emotional data, and normalizes your data so it is ethically sound and compliant with regulations like GDPR. Unit 112(a): Adaptive Learning Model. This part incorporates adaptive learning technologies and is dynamically updated based on collected classroom data, thereby improving its ability to recognize culturally and context-specific emotional expressions. • A user interface module 114 that provides educators with real-time access to emotion analysis via an intuitive dashboard. The interface displays visualizations of students' emotional trends, engagement, and recommendations for classroom interventions. This module offers support for desktop-based and mobile-compatible applications, catering to the diverse educational settings where instructors teach. 1. AI-powered emotion recognition system: This system can be used in several situations, for example: Real-time emotion recognition: By using integrated cameras in a classroom or online learning platform, the system can capture real-time data on students' facial expressions. Secondly, the images are processed with CNN-based models to determine emotional states using the emotion recognition unit 102. 2. Face and emotion recognition: Real-time processing unit 104 to extract facial features and categorize detected emotions into predefined categories. Advanced filtering techniques are used to reduce noise and improve the reliability of recognition. 3. Learning platform integration: The LMS integration unit 106 sends classified emotional signals to corresponding educational platforms to ensure real-time synchronization. 4. Lesson adaptations based on emotion recognition: The Personalization Engine 108 makes real-time adjustments to learning materials, tools, and engagement techniques in response to perceived emotional states. For example, if the system detects frustration, it can introduce interactive elements to rekindle student interest. 5. Rapid feedback is designed as an intervention: The Instant Feedback Mechanism 110 creates a feedback loop regarding the emotional state of the student and also for the teachers, so that they can receive warnings regarding the mental health of the students and intervene. 6. Lifelong Learning and Adaptation: The data storage and adaptive learning model of Unit 112 improves their emotional sensitivity by incorporating updates from new lesson details and gradually adapting the experience to accuracy, culture, and language.

[0019] In one embodiment, the emotion recognition unit 102 can be used both in the classroom and in online learning environments. It utilizes facial recognition technology via webcam-enabled devices, making it a generally accessible solution on educational platforms. Furthermore, the personalization engine 108 is configured for multilingual learning environments and adapts teaching strategies in real time to the students' linguistic backgrounds. Using natural language processing (NLP), the system can recognize the specific emotional aspects of a person's language, thus enabling broader interaction across various school settings.

[0020] The LMS Integration Unit 106, as implemented in another embodiment, could enable direct synchronization with third-party smart learning systems. This functionality guarantees support for popular platforms, including Moodle, Blackboard, Google Classroom, and the UAE Ministry of Education's smart learning platforms. In an alternative embodiment, the Feedback Mechanism Unit 110(a) uses predictive analytics to forecast the development of student engagement over time. This feature allows teachers to identify learning difficulties before students lose interest.

[0021] The data storage and adaptive learning model unit 112, for example, is based on federated learning methods that enable the training of AI models while maintaining privacy. It allows institutions to improve the accuracy of emotion recognition without compromising student data security.

[0022] Advantages of the AI-driven emotion recognition system: The current invention offers numerous advantages over existing adaptive learning methods: Instant emotional awareness: The system offers immediate access to emotional data, allowing teachers to spontaneously adapt their teaching methods. Adaptive, AI-driven learning adjustments enable individualized learning combined with real-time capture of students' emotional state data. Easy implementation in educational systems: Compatible with LMS systems and virtual learning environments.

[0023] Scalable for various educational institutions: The invention can be scaled for different application areas, from primary school to higher education and corporate training programs. Implementation of ethical AI: The system is privacy-conscious and ensures compliance with UAE AI and data protection regulations. At the same time, it provides valuable emotional insights. Cultural adaptation: This model is trained to recognize regional emotion associations and implements an emotion classification framework suitable for students in the UAE.

[0024] A revolutionary new system for educational technology: the AI-powered emotion recognition system for adaptive learning. This invention revolutionizes passive learning environments with emotion-aware, adaptive learning frameworks, integrating deep-learning-based emotion recognition with dynamic teaching methods. Because it is adaptable, considers learner needs, and can adjust pedagogical strategies based on real-time emotional input, it achieves higher engagement and better results in a more personalized and inclusive learning context.

[0025] This system reflects the UAE's ambitions for AI-driven, intelligent education and offers a data-centric, scalable solution to maximize academic performance and emotional well-being in modern classrooms.

Claims

[1] An AI-driven emotion recognition system for learning adaptation in educational institutions, comprising: • an emotion recognition unit that tracks students’ facial expressions in real time and classifies the resulting expressions into a specific emotional class from a given sentence through a Convolutional Neural Network (CNN) trained on a large-scale emotion dataset; • a real-time processing unit that extracts facial features and dynamic emotional states of students through adaptive filtering algorithms; • an LMS integration unit adapted to connect discovered emotional knowledge with online and offline learning environments for customizable learning adaptations; • a personalization engine programmed to intelligently adapt training content, response approaches and training materials according to the identified emotional states so that each learner receives individually tailored and more intensive learning experiences; • a feedback loop that sends educators real-time alerts as well as recommended videos to address emotional feedback or teaching gaps, allowing teachers to address negative signals in the moment (e.g., frustrated face, etc.); • AI data storage and adaptive learning model unit that stores anonymized emotion recognition data, updates the AI ​​model and improves it through learning over time, and ensures that the information used is used in compliance with data protection regulations; • a user interface module specifically tailored for educators with a visual overview of student emotional patterns and real-time analysis of their emotions and engagement as the basis for key decisions that lead to improving pedagogical processes in the classroom. [2] The system of claim 1, wherein the emotion recognition unit trains anger, disgust, fear, happiness, sadness, surprise, and neutral emotions by deep learning-based facial expression recognition (FER) with FER2013 or similar datasets. [3] The system of claim 1, wherein the real-time processing unit uses feature extraction techniques and noise reduction filters to improve the accuracy of emotion classification. [4] The system according to claim 1, wherein the LMS integration unit is compatible with (smart) educational platforms, in particular for example but not limited to: Moodle, Blackboard, Google Classroom, Smart Learning Platforms of the Ministry of Education of the United Arab Emirates. [5] The system of claim 1, wherein the personalization engine may be educational content through customizations that include changes in difficulty, pace of instruction, or engagement activities that respond to real-time emotional feedback from students. [6] The system of claim 1, wherein the feedback mechanism uses predictive analytics to anticipate long-term emotional engagement patterns in learners and suggest actionable pedagogical changes before a decline in engagement occurs. [7] The system of claim 1, wherein the data storage and adaptive learning model unit applies federated learning techniques to gradually improve emotion recognition models without compromising student privacy and data security. [8] The system of claim 1, wherein the user interface module provides a user-friendly dashboard accessible via both desktop and mobile applications, displaying emotional patterns and adaptive learning suggestions in real time. [9] The system of claim 1, wherein the AI-driven emotion recognition system leverages multilingual and multicultural learning environments in the UAE context by training emotion recognition models on region-specific expressions applicable to the diverse student population of the UAE. [10] The system of claim 1, wherein the AI-driven emotion recognition system is integrated and deployed in an online education environment, enabling real-time monitoring and feedback in both physical and virtual classrooms and making it highly scalable and versatile in different learning scenarios.

Images