Generative Adversarial Networkにおけるラベル操作による触覚振動情報の変化

抄録

In recent years, many tactile displays have been developed. Haptic displays based on tactile information obtained from actual textures have high reproducibility, and these displays enable advanced virtual reality experiences. However, in order to extend these contents, new tactile information must be collected from actual textures. However, in order to extend these contents, new tactile information must be collected from actual textures, and it is impractical to collect the necessary tactile information one by one. To solve this problem, we propose a method to generate uncollected tactile information from existing tactile information using Generative Adversarial Network (GAN). We generated and evaluated new haptic vibration information by manipulating the label information for generating haptic vibration information with GAN based on the triaxial acceleration information when touching the actual texture.