Galvanic Vestibular Stimulation in Latent Space

2026-07-29

AI Summary

This study addresses the challenge of synthesizing perceptually consistent galvanic vestibular stimulation (GVS) waveforms from semantic textual descriptions to enable programmable embodied feedback. The authors construct the first dataset comprising 100 GVS waveforms paired with 1,526 free-text perceptual descriptions and propose a retrieval-guided one-dimensional convolutional variational autoencoder to establish a semantic mapping from text to GVS waveforms. This work pioneers the linkage between GVS waveforms and natural language descriptions, demonstrating the feasibility of text-conditioned GVS synthesis. Behavioral experiments reveal that participants achieve a 63.33% accuracy (d′ = 0.70, p < 0.001) in identifying semantically congruent waveform–visual pairings, significantly above chance level, thereby confirming that the generated waveforms exhibit discernible semantic perceptual consistency.

Abstract

Galvanic vestibular stimulation (GVS) is widely used to modulate self-orientation, balance, and motion perception; the discriminability of frequency-encoded cues further suggests its potential as a standalone modality for embodied feedback. However, synthesizing GVS waveforms congruent with target events or bodily states remains challenging. GVS waveforms combine current direction, intensity, duration, and onset and offset transitions, yet how these parameters jointly shape users' perceptual and associative responses remains underexplored. To address this gap, we contribute a dataset linking GVS waveforms to free-form experience descriptions, as well as a retrieval-guided generative model for synthesizing candidate waveforms from target descriptions. The dataset comprises 100 GVS waveforms and 1,526 valid free-form sensation descriptions collected from 16 participants. Semantic analysis revealed diverse motion- and force-related sensations, localized bodily sensations, and situational associations. Compared with a participant-preserving permutation baseline, descriptions elicited by the same waveform covered fewer semantic categories (8.18 vs. 9.45) and exhibited a higher dominant-category proportion (26.97% vs. 21.25%; both P < 0.001). Building on this dataset, we implemented the generative model as a retrieval-guided one-dimensional convolutional variational autoencoder. An independent behavioral study recruited 10 participants who had not contributed to the dataset collection. Performance in discriminating congruent from incongruent waveform-visual cue pairings was significantly above chance, with an accuracy of 63.33%, d-prime = 0.70, and p < 0.001. Together, these findings demonstrate the feasibility of text-conditioned GVS synthesis and support the development of GVS as a programmable modality for semantically congruent embodied feedback across interactive scenarios.

Problem

• Galvanic Vestibular Stimulation
• waveform synthesis
• embodied feedback
• perceptual response
• semantic congruence

Innovation

• Galvanic Vestibular Stimulation
• text-conditioned synthesis
• retrieval-guided generative model
• embodied feedback
• variational autoencoder