[4O5-IS-2c] Machine learning

16:15 〜 16:30

[4O5-IS-2c-04] Personal health prediction and recovery support using evolutionary AI: A new approach integrating forest environment analysis with satellite data

〇Tomoki Furukawa¹, Ogura Hikari¹, Daichi Yamanouchi¹, Nonoka Ooba¹, MASARU IWAKI¹, KIYOMI ASAI¹ (1. IWASAKI GAKUEN)

work-in-progress

キーワード：

Physical condition prediction、Physical condition monitoring、AI

This study proposes a novel method for estimating an individual’s health condition by integrating three modalities: vocal features derived from the RAVDESS dataset, facial expression recognition, and gait analysis using PoseNet-based posture estimation. In parallel, the health status of forest environments was visualized through analyses of NDVI and brightness values obtained from Sentinel-2 satellite imagery. By combining personal health indices with environmental assessments, we developed an advanced AI driven support system capable of recommending optimal forest bathing locations tailored to an individual’s condition. This integrated framework demonstrates the potential of multimodal human sensing and remote environmental monitoring to enhance personalized well being interventions.