1. Introduction

In recent years, with the rapid popularization of remote learning, primarily driven by COVID-19, the need for emotional support in virtual learning environments has gradually emerged. Students need emotional support and guidance on cognitive skills during remote learning to maintain motivation and control emotions for isolation or stress. Affective AI has emerged as an innovative method for identifying and reacting to a student's emotional state, as this transition has driven advances in educational technology. Affective AI technology uses facial expression recognition, speech sentiment analysis and other methods to capture students' emotional fluctuations during the learning process and provide personalized feedback to help students adjust their emotions and improve learning outcomes. In contrast with traditional educational technology, affective AI technology emphasizes students' emotional well-being and enhances their mental abilities, thereby enhancing the learner's overall experience.

Although affective AI technology shows significant potential, existing research focuses mainly on technical implementation. For example, a large body of literature explores how algorithms and sensors can be used to identify students' emotions. However, there is less discussion about the effectiveness of its application in actual teaching scenarios, especially in the specific practice of remote learning environments [1]. Existing research often ignores the ethical and privacy issues facing affective AI technology in practical applications. For example, the collection, storage, and use of affective data may pose risks of privacy disclosure and data abuse. In addition, the collection of affective data not only involves privacy issues but may also lead to unfairness due to the bias of affective recognition algorithms, which differences in gender, race, or cultural background may trigger. Whether students can choose to participate in affective monitoring, whether affective recognition will lead to bias, and the applicability of affective AI technology in different cultural contexts are all challenges that have not been adequately discussed in the current literature. The neglect of these issues limits the widespread use of affective AI technology in education and hinders its fundamental role in remote learning environments.

With the global trend of remote learning, there are significant differences in the acceptance and adaptability of affective AI technology in different cultural backgrounds and learning environments. For example, students from other cultures may express emotions in various ways. This means that affective AI technology may have recognition biases in some cultural contexts, which affects its application in the global education environment. This issue has not been systematically analysed in the existing literature, which poses a challenge to the worldwide promotion of affective AI technology. Therefore, exploring the performance and adaptability of affective AI technology in different cultural contexts has become an important topic that needs to be urgently addressed.

Based on an analysis and review of existing literature, this study aims to examine the current application of affective AI technology in remote learning, with a particular emphasis on its potential ethical and privacy concerns. This study will also explore the adaptability of this technology in different cultural contexts, with a view to providing valuable suggestions and references for the promotion and application of affective AI technology in remote learning around the world. This study will address the research gaps in the literature, identify the significant issues and challenges encountered by affective AI technology in remote learning, and propose corresponding solutions to encourage the further development of future research by systematically combing through existing research.

2. The Theoretical Framework of Emotional AI Technology in Education

In recent years, we have seen an increase in affective AI technology, an emerging educational tool. Personalized support and feedback are provided by recognizing and evaluating learners' emotions in real time, employing the principles of emotional computing. In contrast to conventional educational technology, affective AI emphasizes cognitive development and provides emotional support that helps students manage the learning process's stress and challenges, enhancing learning outcomes. Examining the theoretical foundations and evaluating their specific applications in education is essential for understanding the function of affective AI technology and its potential.

2.1. Application of Affective Computing Theory in Education

The concept of affective computing was first introduced by Professor Rosalind Picard of the MIT Media Lab in 1997. The purpose of this concept was to make it possible computers to recognize, interpret, and respond to human emotions, using computational methods [2]. The theory's basic mechanism includes the collection of the user's mental and emotional state through a variety of physiological signals, such as facial expressions, vocal attributes, heart rate, and skin conductance, so as to enhance the efficacy and simplicity of human-computer interaction. Following that, algorithms have been carried out to generate customized feedback from this data. The theoretical basis for affective AI technologies is determined in educational environments through affective computing, helping teachers and learning systems to identify students' emotions, thereby enhancing the learning experience.

Affective computing is particularly vital in remote learning scenarios, as it is challenging for teachers to perceive students' nonverbal signals through the screen. Affective computing techniques enable teachers or systems to modify their instructional approaches by tracking students' emotional variations in real time [3]. For instance, when the system identifies a student's confusion or anxiety, it can modify the content's difficulty or offer supplementary instruction to assist the student in overcoming emotional obstacles, thereby sustaining motivation and concentration. Affective computing emphasizes the recognition of emotions and includes emotion regulation and feedback systems. The system uses affective computing to automatically modify the difficulty of learning tasks according to students' emotions, ensuring they are in an optimal emotional state for learning. The system can either increase the complexity of a learning task or support a student experiencing distress to try to help them regain focus. Consequently, affective computing provides a theoretical foundation for the broad use of affective AI technology in education, demonstrating substantial potential by providing emotional feedback and regulation systems to improve personalized learning experiences, particularly in remote learning scenarios.

2.2. Educational Significance of Emotion Recognition and Response Technology

Affective computing is a field that uses multimodal perception systems, such as facial expressions, vocal emotions, and behavioral analysis, to identify students' emotions. Affect recognition technology is a critical component of this field [4]. Emotion recognition technology uses machine learning algorithms to analyze students' physiological signals in real time and provide personalized learning support. Emotion recognition technology has the potential to assist teachers or learning systems in real-time in adjusting their teaching strategies by detecting students' micro-emotional fluctuations, compared to the one-way interaction in traditional teaching.

A fundamental benefit of emotion recognition technology in education is the ability to deliver immediate feedback. For instance, when the system identifies a student displaying confusion or decreased focus, it can assist the student in overcoming the existing learning barrier by simplifying the task or offering supplementary explanations. This prompt feedback improves students' learning efficacy and increases the interactivity of the educational process. Utilizing this affective feedback system, teachers can assess students' emotional conditions and modify their instructional approaches to guarantee an optimal learning experience. Moreover, emotion recognition technology extends beyond the detection of transient emotional variations, it is capable of discerning students' ongoing emotional patterns through the collection of emotional data [5]. The system can determine how much a student is experiencing ongoing anxiety or exhaustion as a result of this continuous affective data analysis. Therefore, the system provides affective support tools, such as personalized learning path modifications or emotion regulation strategies. This support system, which is based on long-term emotional trends, can help students manage their emotions and improve their self-efficacy when they face challenges with learning more effectively. Emotion recognition technology can also comprehensively capture students' affective states by combining multiple perceptual signals, providing data support for individualized learning. As a result, emotion recognition technology, a critical component of affective computing, provides a strong technological foundation for remote learning, personalized learning, and emotional support in education.

2.3. Affective Support Models in Remote Learning

Affective support models are essential in remote learning environments. The elements of an effective support model typically include emotion recognition, emotion feedback, emotion regulation systems, and long-term emotion data analysis. Emotion recognition uses a multimodal perception system to identify students' emotional states and deliver specific feedback. Informed of this input, the system can offer students emotional support through a dynamic adjustment system to help them handle the learning challenges of emotional fluctuations. The affective support model initially recognizes students' emotional changes throughout the learning process via regular emotional evaluations and subsequently offers personalized emotional feedback and guidance. After detecting students' learning burnout or emotional fatigue, the system may provide specific emotional regulation strategies.

The affective support model effectively enhances students' learning motivation by introducing incentives. The system can offer suitable emotional support based on the student's emotional condition and assist them in regaining their interest in learning during periods of depression. This incentive mechanism may reduce learning disruptions resulting from emotional fluctuations and provide a more consistent emotional experience for students in remote education. Meanwhile, the affective support model emphasizes the importance of students' self-regulation abilities. The model offers self-regulation tools to assist students in managing their emotions under learning pressure. The tools include evaluation and feedback on emotional conditions and targeted operational recommendations for emotion management, such as relaxation training and emotion regulation techniques. By utilizing self-regulation tools, students can gradually improve their emotional management abilities throughout the learning process, sustaining emotional stability and motivation when confronted with prolonged studies.

Studies indicate that implementing affective support modeling significantly enhances students' learning outcomes and their capacity to manage emotional difficulties [6]. By continuously regulating and managing their emotions, students may maintain motivation and continuity in learning, reduce burnout through emotional fluctuations, and substantially improve overall learning performance and satisfaction. Therefore, the role of affective support modeling in remote learning must be acknowledged to ensure that students receive ongoing emotional assistance when facing challenges.

3. Prospects of Emotional AI Technology in Remote Learning

With the continuous development of affective AI technology, its application potential in remote learning is becoming increasingly prominent. Affective AI technology helps students maintain a high level of learning motivation and improves their learning outcomes through personalized learning path optimization. In addition, affective AI technology provides a new perspective on student mental health support. In the context of remote learning, the application prospects of affective AI technology are broad and far-reaching.

3.1. Potential Role in Enhancing Learning Motivation and Outcomes

In remote learning environments, students' academic success depends on their motivation. Through real-time emotion monitoring and feedback, affective AI technology can help students sustain a positive learning environment and increase their learning motivation. Research suggests that students' motivation increases when they are introduced to positive emotions, which enhances their ability to focus on educational tasks and their understanding and knowledge of the material [7]. The affective AI system has the potential to discover a student's emotional state by analyzing their facial expressions and vocal intonations, as well as providing personalized instruction to foster long-term motivation.

Affective AI technology enhances learning by adaptively modifying the complexity of educational tasks. When detecting a student's anxiety, confusion, or frustration, the system simplifies the task's complexity. It provides enhanced instructional support to help students overcome challenges and regain confidence in their studies. Conversely, when a student experiences a more favorable emotional state, the system increases the learning challenge, motivating the student to achieve ongoing advancement. Therefore, affective AI technology enhances student motivation and significantly improves academic performance and learning efficiency.

Remote learning lacks face-to-face emotional communication, and affective AI can bridge this gap by offering timely emotional support to students. Continuous emotional feedback enables students to maintain increased engagement in learning and reduce disruptions caused by emotional fluctuations. This emotional support is particularly advantageous for students in remote learning settings who lack the immediate teacher-student interaction typical in traditional classrooms.

3.2. Development and Optimization of Personalized Learning Paths

Personalized learning represents a crucial trajectory in modern educational technology, while affective AI technology provides new possibilities for implementing tailored learning experiences. Affective AI technology can dynamically modify learning content and pathways in response to students' moods through real-time emotion recognition. This technology enables the system to deliver a personalized learning experience tailored to each student's educational requirements and emotional condition, facilitating a more flexible and adaptable learning process.

Traditional remote education often depends on standardized course schedules and materials, overlooking the unique requirements of students. By continuously monitoring students' emotional responses, affective AI technology promotes personalized learning paths. For example, the system may reduce the complexity of the task or offer supplementary learning resources to minimize anxiety and accelerate the student's return to a productive learning state when it recognizes a student's emotional state as indicative of fatigue or nervousness. To preserve an engaging learning experience, the system may increase the task's difficulty or introduce new learning challenges when the student displays positive emotions.

Affective AI technology can identify students' learning patterns and emotional trends through extensive data collection to enhance personalized learning pathways [8]. Utilizing this emotional data, the system may help students gradually improve their self-regulation skills to effectively manage learning challenges during various emotional states. Optimizing personalized learning paths enhances educational outcomes, renders remote learning more student-centric, and offers more effective learning assistance.

3.3. Prospects for Supporting Students' Mental Health

Students' mental health concerns have grown more prevalent as remote learning becomes increasingly popular. In the remote learning environment, students' academic performance is negatively impacted by negative emotions such as anxiety, stress, and loneliness [9]. However, it is challenging to address these problems in time using traditional teaching techniques. Affective AI technology may provide students deeper affective support by providing real-time monitoring and feedback, which allows them to more effectively manage the psychological challenges of the learning process. By identifying students' emotions, affective AI technology can assist teachers and students in identifying mental health issues in early stages. For instance, the emotional AI technology can generate alerts to remind teachers or parents to monitor changes in student emotions when students display signs such as anxiety, depression, or fatigue during the learning process. The emotional AI system can also assist students in reducing mental strain through its emotional regulation functions, which include providing supportive feedback or recommending that students take breaks as needed while creating necessary adjustments.

Affective AI technology can enhance students' learning mindset and support mental health. It utilizes emotional feedback to help students maintain emotional stability and prevent disruptions to learning induced by negative emotions. Studies indicate that affective AI may assist students in maintaining a positive emotional state through frustration by analyzing their feelings in real time, reducing anxiety [6]. This shows that affective AI technology has diverse applications in supporting students' mental health, particularly in remote learning scenarios, where it can help their mental and emotional wellness.

4. Challenges of Emotional AI Technology in Remote Learning

Affective AI technology shows significant potential in remote learning; however, its application faces many obstacles. Before it can be broadly implemented, affective AI technology must address several practical challenges, particularly concerning data privacy, ethical considerations, and user acceptance. Developing privacy protection methods, refining algorithm design, increasing cultural adaptability, and promoting technology acceptance through user education can more effectively actualize the potential of affective AI technology.

4.1. Technical Challenges and Algorithm Limitations

Affective AI technology analyzes and provides perspectives on learners' emotional conditions by gathering students' affective data, posing a risk of privacy breaches. Implementing affective AI technology will involve recording, storing, and evaluating significant quantities of students' affective data [10]. This data is frequently susceptible and can include information regarding personal emotional states, mental health, and additional factors. The leakage or misuse of information will significantly jeopardize students' privacy and security. To resolve this issue, it is imperative to enhance data encryption and anonymization technologies. Encryption technology may prevent unauthorized access and ensure data security during transmission and storage. Moreover, anonymization ensures data is analyzed post-removal of personally identifiable information, avoiding direct association with particular individuals. Crucially, educational platforms should adhere to minimal data collection, gathering only essential sentiment data to mitigate the risk of data breaches.

From a legal perspective, regulations such as the General Data Protection Regulation set explicit criteria for processing affective data, especially concerning informed consent and data transparency. Educational platforms must explicitly communicate to students about using their affective data and protect their personal rights throughout data collection and processing [11]. However, with the development of affective AI technology, technical methods and legal frameworks for privacy protection still need to be constantly updated to meet the increasingly complex needs of affective data protection.

4.2. Issues of Data Ethics and Fairness Issues

Affective AI technology prompts numerous inquiries regarding ethical considerations and fairness in its implementation. The collection of affective data necessitates continuous monitoring of students' emotional states through devices like cameras and microphones, potentially intruding upon students' autonomy and inducing psychological distress [12]. Students must possess the unequivocal right to information and independence, enabling them to determine their participation in affective monitoring and to maintain sufficient understanding and control over the use of affective data. In addition, continuous dependence on affective AI for emotional control could reduce students' ability to manage their emotions. Students who overly depend on the feedback and assistance of affective AI may impair their emotional regulation, negatively impacting their emotional maturity and ability to manage challenges. Therefore, it is imperative to balance the implementation of affective AI with the cultivation of students' emotional autonomy in the design of such systems.

The use of affective AI technology may also influence the evaluation of teaching. Excessive dependence on emotion analysis provided by affective AI may result in the inability to detect students' real learning abilities when emotional data is used for learning evaluation. Affective AI can identify changes in students' emotions and supply immediate feedback, but this emotional data shouldn't be used as the sole basis for student evaluation. For example, certain students exhibit increased levels of anxiety or emotional instability when faced with obstacles; however, this is not always a sign that their educational achievement will be impacted. Similarly, affective AI technology may place more emphasis on outward emotions in its analysis while ignoring the actual learning outcomes of some introverted students or students who are not good at expressing emotions. Therefore, a balanced approach is necessary in the use of affective AI technology, ensuring it serves as a supplementary tool in conjunction with other forms of learning evaluation. This approach guarantees the fairness and comprehensiveness of the evaluation system.

4.3. User Acceptance and Cultural Diversity Issues

The global application of affective AI technology is confronted with challenges related to cultural diversity and user acceptance. This presents adaptation challenges for implementing affective AI systems in diverse environments, as students from various cultures express their emotions in an individual way. For example, students from specific cultures may be unable to express their internal emotions through blatant expressions, which could lead to the mistakes of emotion recognition in affective AI technology if it fails to accurately identify their actual emotional states [13]. The effectiveness of affective AI systems in remote learning may be reduced due to recognition bias that may emerge because of cultural disparities across various contexts. The system design must exhibit significant cultural sensitivity to improve the adaptability and user acceptance of affective AI technology across various cultural contexts. The affective AI system must comprehend and adjust to the variations in emotional expression across cultures and offer multiple choices in the emotional feedback system. The algorithm must adaptively modify the parameters and feedback systems of emotion recognition based on the emotional expression patterns of students from diverse cultures to ensure that they receive accurate and suitable affective support within their cultural environment. Advanced AI technology may achieve broader global applications by enhancing adaptability.

The promotion and user acceptance of affective AI technology is contingent upon users' understanding and trust in the technology's functionality and privacy protections. A sufficient knowledge of the affective AI system's operational principles and security features among students and teachers may lead to resistance, hindering its practical implementation. Therefore, when pushing for affective AI technology, educational platforms must assist students and educators in comprehending the technology's proper function and boost their confidence in privacy safeguards through thorough user education and training. Educational platforms may alleviate users' concerns and enhance their acceptance of technology by providing clear privacy protection protocols and transparency measures regarding emotional data processing.

5. Conclusion

This study examines the opportunities and obstacles associated with introducing affective AI technologies in remote learning, emphasizing their potential to improve learning motivation, personalize individual educational paths, and promote students' mental well-being. The primary benefit of these technologies is their ability to provide real-time monitoring and feedback on students' emotional states, improving education's personalization and interactivity. However, the broad use of affective AI technology is met with significant challenges, including the safety of data privacy, the fairness of learning, ethical considerations, and cultural adaptability. In particular, the algorithmic bias of affective recognition and the privacy protection of affective data may have a detrimental effect on the learning experience and rights of students. Therefore, it is imperative to implement additional technical improvements and regulations.

The possibility of applying affective AI technology is extensive; however, it necessitates resolving many practical challenges and obstacles. Developing more advanced technical solutions for privacy protection and fairness in future research has become imperative. At the same time, enhancing the flexibility of affective recognition in various cultural settings is essential. Furthermore, the effectiveness of user confidence and acceptance of affective AI technology can be increased through improved technical transparency and user education. These enhancements guarantee that AI technology contributes to remote learning and provides ongoing innovation in education systems worldwide.

Future research should maintain its emphasis on improving affective AI technology in terms of cultural diversity and ethics. Investigating its practical applications in long-term learning scenarios is crucial in this context. Affective AI technology is expected to significantly impact future educational environments and drive the further development and transformation of remote learning by systematically addressing these challenges.