Optimization of Sound of Autonomous Sensory Meridian Response with Interactive Genetic Algorithm

Movie contents of Autonomous Sensory Meridian Response (ASMR) are popular among Internet users. The users enjoy positive feelings caused by the binaural sounds via headphones. Some previous studies clarified the psychophysiological effects of the ASMR contents. Referring to a previous proposal for creating ASMR sounds by Miyazaki et al., this study proposes an Interactive Evolutionary Computation for creating ASMR sounds. Interactive Evolutionary Computation has been applied for creating media contents including sound and music contents. The property of the proposed Interactive Evolutionary Computation is creating ASMR sounds suited to each user's preference and feelings by simple repetitive evaluations. For having a fundamental investigation of the efficiency of the proposed method, a listening experiment composed of two steps was conducted. Each of the twelve subjects participated in both of the two steps, and they evaluated the ASMR sounds and scored them repeatedly in the search step. In the evaluation step, they scored the two sounds picked up from the 1st and the 10th generations. Trend of increase in the fitness values was observed in both of the experimental steps.