Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 3, 2912-2918 2025 Publisher: Learning Gate DOI: 10.55214/25768484.v9i3.5887 © 2025 by the author; licensee Learning Gate © 2025 by the author; licensee Learning Gate History: Received: 30 January 2025; Revised: 10 March 2025; Accepted: 14 March 2025; Published: 29 March 2025 * Correspondence: youngho@duksung.ac.kr Emotion recognition and typographic expression in AI media art: A study on sign Young Ho Kim1* 1Department of VR Convergence Engineering, Duksung Women's University, Korea; youngho@duksung.ac.kr (Y.H.K.). Abstract: This study investigates the integration of artificial intelligence (AI) into art, focusing on its influence on emotional expression and human-AI interaction through the interactive artwork, Sign. Utilizing real-time emotion recognition and font psychology, Sign captures participants' facial expressions, analyzes their emotional states, and visually reflects these through dynamic font transformations. By blending technology with creativity, the artwork offers a novel medium for participants to engage with their emotions while highlighting AI's role in expanding artistic expression. This study underscores the innovative potential of AI-art collaborations to reshape emotional communication. Despite challenges such as the cultural variability of emotion representation, the findings present pathways for integrating AI to enhance human-AI interactions and emotional experiences in artistic contexts. Keywords: Artificial intelligence, Emotion recognition, Font psychology, Interaction, Interactive media art. 1. Introduction In contemporary society, advancements in artificial intelligence (AI) technology are reshaping the relationship between technology and art, profoundly influencing artistic processes and outcomes. AI has moved beyond its role as a tool to become an integral part of the creative process, enabling artists to explore innovative methodologies and expand their creative horizons. This integration enriches artistic works with deeper emotional complexity and diverse expressions. Beyond the creation of novel artistic forms, AI facilitates new modes of audience engagement, offering unique emotional experiences and enhancing the communication of artistic intent. These transformations challenge traditional notions of art, prompting a redefinition of its purpose and value. In this study, the focus is placed on the intersection of AI and art to investigate how AI influences artistic expression and emotional communication through the interactive artwork Sign. By employing AI-driven emotion recognition and dynamic font transformations, Sign captures and visualizes participants’ emotions in real-time. This research delves into the integration of AI technology into artistic creation, exploring the ways it expands the diversity and complexity of emotional expression in human-AI interactions. Furthermore, it examines the synergistic relationship between technology and art, revealing how their convergence can redefine creative processes and artistic outcomes. By unpacking these aspects, the study sheds light on the transformative role of AI in reshaping the future of art and human-technology interaction. 2. Literature Review Recent years have witnessed growing scholarly interest in the intersection of artificial intelligence (AI) and artistic expression, particularly in the realms of emotion recognition and its applications. Foundational studies in emotion recognition have established key theories supporting the integration of emotional signals into human-machine interactions, emphasizing that accurate emotion decoding mailto:youngho@duksung.ac.kr 2913 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 3: 2912-2918, 2025 DOI: 10.55214/25768484.v9i3.5887 © 2025 by the author; licensee Learning Gate enhances the depth and quality of these interactions [1]. Research within human-computer interaction underscores the importance of emotional expression, revealing that users exhibit social behaviors during engagements with AI systems, thereby fostering meaningful connections [2]. Practical implementations of AI-driven emotion recognition in healthcare further demonstrate its potential to address real-world challenges, such as improving patient engagement and supporting therapeutic interventions [3]. The development of emotion recognition systems has been underpinned by studies on universal emotional expressions and their interpretative mechanisms, which form the basis for advancing AI’s ability to discern complex emotional states [4]. Investigations into human emotional regulation and decision-making have provided valuable insights into how AI systems can analyze and respond to multifaceted emotional phenomena [5]. In the context of artistic representation, research on font psychology has illuminated the emotional impact of typography, offering frameworks for understanding how visual elements influence emotional perception [6]. Recent advancements in deep learning have enabled real-time emotion visualization, enriching the interactive and immersive experiences of artistic works [7]. Additionally, studies on AI’s role in interactive platforms, such as gaming, highlight the potential for emotion recognition to personalize user experiences and enhance creative applications [8]. These studies collectively highlight the transformative potential of AI in the artistic domain. They provide critical foundations for this study, which examines how the Sign artwork integrates emotion recognition and font psychology to redefine emotional expression and human-AI interaction. 3. Introduction and Meaning of the Sign Artwork 3.1. Concept of the Artwork The Sign artwork embodies the emotional nuances encountered throughout the process of cultivating relationships between artificial intelligence and human beings. In a broader context, a "sign" denotes consensus, wherein distinctive interpretations are articulated through the idiosyncratic handwriting of the signer. While conventionally functioning as a conduit for establishing trust, the Sign artwork amplifies this notion to encompass a variety of emotional articulations, including facial expressions, gestures, and vocal intonations. These manifestations not only illuminate the intricacies of human relationships but also investigate novel avenues for interaction between humanity and artificial intelligence. The Sign artwork transcends conventional interhuman dynamics to probe into the progressive relationship between human beings and technological advancements. 3.2. The Significance of Human Signatures Throughout the annals of history, signatures have constituted an essential mechanism for articulating intent and consensus. They embody individuality and distinctiveness through the unique penmanship of the signer, functioning as corroborative evidence for formal transactions such as contracts and agreements. Originally inscribed on paper or wax utilizing a writing instrument, signatures have undergone a transformation alongside technological advancements, manifesting in diverse forms. Contemporary signatures, encompassing electronic signatures and biometric authentication modalities like fingerprints, transcend their conventional legal purposes and have emerged as instruments for emotional articulation. They amalgamate multifaceted components such as handwriting, facial expressions, gestures, and vocal inflections, acting as pivotal mediators for the communication of human intent and emotions. This progression situates signatures as a medium that not only mirrors personal identity but also elucidates emotional and social affiliations. The Sign artwork investigates the broadened implications of signatures by reinterpreting them as instruments for emotional articulation. The signatures of participants transcend mere acquiescence, vividly representing their emotional conditions through distinctive, dynamically generated forms. This process synthesizes both traditional and contemporary functions of signatures, artistically portraying human emotions, behaviors, and identities. Through this innovative reinterpretation, Sign illustrates the potential of signatures to function as novel emotional connectors between humans and artificial 2914 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 3: 2912-2918, 2025 DOI: 10.55214/25768484.v9i3.5887 © 2025 by the author; licensee Learning Gate intelligence. 3.3. The Relationship Between Humans and AI in the Artwork In the Sign artwork, participants read a brief textual excerpt and provide a signature, initiating an exploration into human-AI interaction. The AI system uses facial recognition to analyze emotional states and monitor participants’ responses, transcending the act of signing into a deeper documentation of behaviors and emotions. The emotional analysis results are applied to the signature through font psychology, transforming it into a visual expression of the participant’s emotions. Specific emotional states directly influence the shape and style of the signature, allowing participants to see their emotions reflected in dynamically altered typography. The AI systematically captures and records emotional signals, providing participants with a visual representation of their emotions and encouraging reflection on the connection between their emotional states and their signatures. By integrating emotion analysis into artistic creation, Sign facilitates a deeper understanding of human-AI interaction and its broader implications for technological advancements. Through Sign, participants observe the evolution of human-AI relationships, gaining insight into how AI interprets human emotions and behaviors. The artwork explores new possibilities for emotional connection between humans and technology, presenting this theme in an evocative and artistic manner. Ultimately, Sign broadens our understanding of how humans and technology can coexist and evolve within future societal frameworks. 4. Technical Implementation and Applications of the Sign Artwork 4.1. Application of Emotion Recognition and Analysis Technology The Sign artwork integrates artificial intelligence to analyze participants’ facial expressions and emotional states. Utilizing deep learning and computer vision techniques, the system relies on a Convolutional Neural Network (CNN) to identify facial features and classify a range of emotional expressions. CNN, widely recognized for its precision in image analysis, plays a pivotal role in capturing and interpreting participants' subtle facial cues. Figure 1. Emotion analysis workflow of sign artwork. As illustrated in [Figure 1], the emotion recognition process in Sign consists of four main phases. The first phase, facial detection, identifies participants’ faces and extracts key facial features. This is achieved through traditional methods like Haar Cascade and advanced techniques such as Multi-Task Cascaded Convolutional Networks (MTCNN). MTCNN efficiently maps facial contours and identifies critical regions such as the eyes, nose, and mouth, ensuring robust performance under diverse conditions, including varying lighting, angles, and backgrounds. In the second phase, facial region extraction, the system isolates and processes relevant facial areas to refine data for emotion analysis. This preprocessing step minimizes noise, enhancing the accuracy of subsequent analysis. The third phase, emotion analysis, employs CNN alongside auxiliary algorithms like Support Vector Machine (SVM), Random Forest (RF), and Decision Tree (DT) to classify emotions with precision. When overlapping emotions, such as "happiness" and "surprise," are detected, the system quantifies their relative intensities for multi-dimensional categorization. Finally, the results visualization phase presents emotion analysis outcomes in real-time, using textual and graphical representations. As shown in [Figure 2], this interface, accessible in administrator mode, ensures the system operates effectively. The 2915 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 3: 2912-2918, 2025 DOI: 10.55214/25768484.v9i3.5887 © 2025 by the author; licensee Learning Gate most prominent emotions detected are displayed graphically and influence the dynamic font design based on participants' emotional states. Figure 2. Facial recognition and emotion analysis interface. 4.2. Emotion Visualization Using Font Psychology The Sign artwork underscores the importance of articulating emotions through visual means by altering font styles and shapes predicated on the outcomes of emotion analysis. The alteration of fonts, which constitutes a fundamental aspect of the artwork, serves to dynamically mirror the emotional states of participants as manifested in the text presented on the screen. This mechanism facilitates participants’ direct engagement with the manner in which their emotions are transmuted into visual portrayals. Empirical research has demonstrated that various font styles can invoke distinct emotional responses. For example, rounded and uniform fonts are associated with sensations of stability and happiness, whereas angular fonts characterized by sharp lines communicate feelings of anger and tension. Drawing upon such empirical findings, the Sign artwork selectively and dynamically chooses and implements fonts that correspond to particular emotional states. In instances where happiness is detected, the font transitions into a rounded and warm design, symbolizing positivity. In contrast, when sadness is recognized, the font transforms into a slender and slanted style with increased spacing, signifying solitude and introspection. For visceral emotions such as anger, the font adopts bold and angular configurations to visually depict heightened energy. These dynamic transformations extend beyond mere aesthetic considerations, functioning as instruments for participants to investigate and comprehend their emotions through an innovative visual medium. 2916 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 3: 2912-2918, 2025 DOI: 10.55214/25768484.v9i3.5887 © 2025 by the author; licensee Learning Gate Table 1. Emotion and font characteristics. Emotion Font Characteristics Description Neutral Simple, even spacing, sans-serif Reflects balance and objectivity. Clean and minimal design emphasizes neutrality and clarity. Happiness Rounded shapes, vibrant colors, smooth curves Conveys positivity and joy. Soft curves and lively colors enhance a cheerful and inviting atmosphere. Surprise Bold, dynamic changes in size and spacing, playful shapes Suggests astonishment and excitement. Abrupt variations mirror the feeling of unpredictability. Sadness Thin, serif, italicized fonts, wide letter spacing Reflects introspection and melancholy. Gentle strokes evoke a sense of isolation and contemplation. Anger Bold, sharp angles, irregular spacing Conveys intensity and tension. Harsh lines and strong contrasts emphasize frustration and aggression. Disgust Jagged, uneven strokes, muted or clashing colors Suggests aversion and discomfort. Irregular forms and unappealing designs mimic feelings of repulsion. Others Experimental, abstract designs, varied textures Represents ambiguity or complexity. Unique and unconventional fonts are used to convey diverse emotions 5. Final Visualization The Sign artwork functions as an interactive installation centered around a single tablet display. Participants engage with the artwork by reading a short phrase displayed on the tablet and signing their name in the designated area. Once the signature is completed, the system captures and analyzes the participant’s facial expressions in real time to extract emotion-related data. This emotional information is instantly reflected in the artwork through dynamic font transformations, visually expressing the participant’s emotional state in the modified text displayed on the screen. The results of the analysis are presented as simplified emotional metrics alongside the transformed text. For instance, happiness is visualized through rounded and soft fonts, exuding warmth and positivity, while sadness is expressed with thin, slanted typography, symbolizing introspection and solitude. These visual changes offer participants a unique opportunity to reinterpret their emotional states through artistic expression, fostering a deeper connection between their emotions and the creative medium. Figure 3. Audience Interaction with the Artwork. 2917 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 3: 2912-2918, 2025 DOI: 10.55214/25768484.v9i3.5887 © 2025 by the author; licensee Learning Gate Figure 4. Results visualization (Left: Default, right: Transformed appearance). 6. Conclusion This study explored the integration of artificial intelligence (AI) with art through the interactive Sign artwork, focusing on its impact on human-machine interaction and emotional expression. By analyzing facial expressions and dynamically transforming font styles, the artwork redefines artistic boundaries and fosters immersive, emotional experiences. Sign places emotional data at the center of its artistic expression, allowing participants to see their emotions reflected in real-time and deepening their engagement with the piece. By incorporating font psychology, the artwork offers participants a creative lens to reinterpret their emotions visually, bridging human experiences with technological innovation. However, challenges remain, including the current limitations of emotion recognition systems in capturing subtle or complex emotions and accounting for cultural and personal variations in emotional expression. Future developments could refine emotion analysis through advanced deep learning techniques and culturally inclusive datasets. Expanding the system to accommodate multilingual writing and personalized feedback loops would enhance the artwork’s universality and emotional depth. Sign exemplifies the transformative potential of AI in art, paving the way for collaborative expressions of creativity between humans and technology. Transparency: The author confirms that the manuscript is an honest, accurate, and transparent account of the study; that no vital features of the study have been omitted; and that any discrepancies from the study as planned have been explained. This study followed all ethical practices during writing. Copyright: © 2025 by the authors. This open-access article is distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/). References [1] R. W. Picard, Affective computing. Cambridge, MA: MIT Press, 1997. [2] B. Reeves and C. Nass, The media equation: How people treat computers, television, and new media like real people and places. Stanford, CA: CSLI Publications, 1996. [3] L. D. Riek, "Healthcare robotics," Communications of the ACM, vol. 60, no. 11, pp. 68-78, 2017. https://doi.org/10.1145/3127874 [4] P. Ekman, "An argument for basic emotions," Cognition & Emotion, vol. 6, no. 3-4, pp. 169-200, 1992. https://doi.org/10.1080/02699939208411068 [5] C. E. Löckenhoff and L. L. Carstensen, "Socioemotional selectivity theory, aging, and health: The increasingly delicate balance between regulating emotions and making tough choices," Journal of Personality, vol. 72, no. 6, pp. 1395-1424, 2004. https://doi.org/10.1111/j.1467-6494.2004.00301.x https://creativecommons.org/licenses/by/4.0/ https://doi.org/10.1145/3127874 https://doi.org/10.1080/02699939208411068 https://doi.org/10.1111/j.1467-6494.2004.00301.x 2918 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 3: 2912-2918, 2025 DOI: 10.55214/25768484.v9i3.5887 © 2025 by the author; licensee Learning Gate [6] E. R. Brumberger, "The rhetoric of typography: The awareness and impact of typeface appropriateness," Technical Communication, vol. 50, no. 2, pp. 224-231, 2003. [7] H. Fujita, Y. Yoshikawa, and T. Nomura, "Real-time emotion visualization using deep learning for interactive media art," Multimedia Tools and Applications, vol. 79, no. 33–34, pp. 24253–24273, 2020. https://doi.org/10.1007/s11042- 020-09087-w [8] G. N. Yannakakis and J. Togelius, Artificial intelligence and games. Springer. https://doi.org/10.1007/978-3-319- 63519-4, 2018. https://doi.org/10.1007/s11042-020-09087-w https://doi.org/10.1007/s11042-020-09087-w https://doi.org/10.1007/978-3-319-63519-4 https://doi.org/10.1007/978-3-319-63519-4