Visuo-haptic mismatch in VR can induce discomfort similar to the uncanny valley effect. The research systematically manipulates delay and force gaps between vision and touch. Physiological signals (HRV, EMG, skin conductance, pupil dilation) are used to quantify discomfort. The goal is to build a predictive model and design guidelines for emotionally congruent VR systems.
*JSPS Kakenhi Kiban C grant - FY2026 to FY2028Our research uses facial video to estimate a person’s mental and emotional state without requiring physical sensors. From face and eye images, we extract remote photoplethysmography (rPPG) signals, facial expressions, landmarks, and other biometric features. AI models combine these signals to identify patterns related to emotions, stress, fear, engagement, and confidence. This approach can help us understand how people respond during everyday activities, learning, and human–computer interaction. Our goal is to develop comfortable, contactless systems that support education, healthcare, and affect-aware technology.
When we look at the two lines from the telescope, we see the illusion of two lines; instead of being converged at the far end, it reverses the perspective. We have been doing experiments to prove the cause of this visual illusion.
A hybrid image is an image that changes the content by the viewing distance. Our lab has successfully combined two different shape images into a hybrid image (above is a cat and car; try viewing very close or very far to see the other image); we are working on combining more than two images. The image is created by interplaying with the image frequency using the image processing technique. This technique, rooted in vision science, reveals how human perception processes spatial information.
VR enables real-time interaction, allowing clients to explore and give instant feedback on design elements, leading to efficient decision-making and increased satisfaction for both designers and clients.
We investigated audio models designed to reduce overlapping voices and improve speech clarity during VR poster presentations. The work explored how spatial, directional, and area-based audio can improve users’ comfort, focus, and social interaction. Our findings support the development of more immersive and usable audio environments for large-group communication in the Metaverse.