🤖 AI Summary
This study addresses the limitation in text-based AI conversational interfaces where role differentiation relies solely on explicit labels or layout cues, lacking spatial metaphor. We propose an implicit visual role-encoding method leveraging chromostereopsis—the perceptual illusion of depth induced by color wavelength differences. By modeling human visual perception and employing a contrast-driven depth-mapping algorithm, we assign users and AI agents to distinct chromatic depth planes, thereby conveying role hierarchy and physical spatial separation within unmodified plain-text interfaces—without altering structure or adding labels. Implementation uses lightweight CSS/Canvas rendering, ensuring broad compatibility and real-time performance. User studies demonstrate statistically significant improvements in role identification accuracy (+27%, *p* < 0.01) and conversational immersion, with no increase in cognitive load. To our knowledge, this is the first application of chromostereopsis to conversational role design, establishing a novel paradigm for implicit, low-intrusion human–AI role differentiation.
📝 Abstract
We propose leveraging chromostereopsis, a perceptual phenomenon inducing depth perception through color contrast, as a novel approach to visually differentiating conversational roles in text-based AI interfaces. This method aims to implicitly communicate role hierarchy and add a subtle sense of physical space.