Vibrotactile Communications: Delivering Immersive Remote Touch Experiences

Vibrotactile signals play an important role in human perception of haptic interactions and object surfaces. Thus, conveying these cues with high quality is vital for an immersive and pleasant experience. In order to create true immersion, many points of interaction need to be conveyed to human users. Thus, when conveying vibrotactile cues over the internet, a large number of signal channels must be transmitted that results in a high data throughput. In order to enable this, we design efficient data compression methods. For the human user receiving the signal to perceive these signals at high quality, the algorithms must be able to compress them perceptually. Therefore, we design custom and detailed perceptual models that allow the compression algorithm to remove imperceivable signal information and thus compress signal data. Additionally, the received vibrotactile signals can be enhanced in their quality before being rendered, for which we developed a neural network that reconstructs some of the lost signal information. Finally, as the signals are displayed to users, actuators can introduce unwanted distortions. These distortions should be mitigated, for which we have developed nonlinear adaptive filter-based methods.