Lie Detection in Social VR Using Multimodal Data

Trust is a key element in human relationships and is vital for the success of any interaction. Establishing trust in Social VR is important for promoting collaboration, communication, and social interaction. In our research, we investigate the feasibility of using lie detection techniques in virtual reality to build trust. Our findings suggest that it is possible to detect deception in virtual reality using sensor-detected cues. Using random forest, gaze-based features by themselves could reach an 87% accuracy in detecting lies.