Immersive visual-thermal stimulation applied to a third person virtual avatar modulates upper limb skin temperature

Abstract

Immersive VR can evoke a sense of physical presence and realism. While visual-thermal cues on a first-person avatar influence thermal perception and skin temperature, it is unclear whether these effects stem from visual input alone or are mediated by self-identification. We examined this using a motion-mimicking third-person (3P) avatar and a VR mirror to manipulate self-identification. Two scenarios simulated cold or hot stimulation on the avatar’s right arm with matching ambient lighting (blue/red). In the Sync condition, the avatar mirrored participant movement; in Async, it moved independently. Nine participants (5 female, M = 28.2 ± 5.6) took part. Skin temperature was measured at six sites on the arms and hands, and a 10-item questionnaire assessed presence, self-identification, and thermal experience. A reference session without visual input controlled for HMD-related temperature effects. Presence ratings (6.55 ± 0.21) were higher than the neutral rating (5.0). Self-identification scores were significantly higher in Sync than Async across both stimuli (p < .001). Participants perceived visual thermal cues as realistic (6.66 ± 0.37). Skin temperature increased in the cold condition (ΔT = +0.05°C) and decreased in the hot condition (ΔT = –0.11°C), regardless of synchrony. Paired t-tests showed significant differences between cold–sync and hot–sync (p = .023), and between cold–async and hot–async (p = .007), indicating cue polarity effects. Although differences between Sync and Async were observed (cold: ΔT = +0.05°C; hot: ΔT = –0.02°C), they were not statistically significant, likely due to the small sample size. These preliminary results suggest that visual thermal cues in VR can influence both subjective and physiological responses. While the role of self-identification did not reach significance, it remains a promising factor. Further data collection is planned to validate this possibility.