Interactive Emotional Expression through Art and AI
The Recoloring Tool is an interactive system that empowers users to create expressive digital paintings through real-time facial emotion recognition. By analyzing emotional states and modifying color palettes, it offers a personalized and resonant art-making experience. This innovative tool blends AI-driven emotion recognition with artistic creativity, pushing the boundaries of traditional digital art.
As interactive experiences become more emotionally driven, the Recoloring Tool explores how AI and emotional data can be used to create deeper, more personalized player experiences in experimental games and interactive art. This project presents new ways to engage users emotionally, breaking away from traditional game mechanics to offer emotionally-driven gameplay mechanics.
Through the use of affective computing, this project opens possibilities for designing interactive systems that adapt not just to player actions but to their emotional states, potentially inspiring new approaches in game design where players’ emotions shape the narrative and gameplay experience.
Key Features:
Emotion-Aware Interaction: The tool captures users’ facial expressions, translating them into color changes, creating a dynamic link between the user’s emotions and their creative output.
Real-Time Feedback: Users see instant changes in their artwork as their emotions shift, allowing for real-time engagement with their creative process.
Art and Technology Fusion: This project exemplifies the seamless integration of artistic creativity with cutting-edge technology, demonstrating the potential of emotion-aware tools in artistic expression.
This schematic diagram illustrates the architecture of our computational tool designed for recoloring artwork based on the user's emotional state. The system integrates facial emotion recognition and a user-friendly interface to facilitate emotional expression through color.
A GUI allows users to choose images from the WikiArt dataset (Tan et al. 2019).
For the automatic palette selection, we used Chang et al.’s algorithm (Chang et al. 2015). By choosing four images (one for each emotion class), a user defines the color combinations that they think correspond to their emotions.
• Through development and evaluation of a recoloring tool, we demonstrated that the tool is helpful for producing expressive paintings and interacting with them using facial expressions.
• We believe that using deep learning technology and stroke-based rendering to build a computational tool opens new ways for users to express their emotions.