@article{oai:kansai-u.repo.nii.ac.jp:02001441, author = {Mae, Yasushi and 前, 泰志 and Nagata, Akihisa and 永田, 暁久 and Takahashi, Tomokazu and 高橋, 智一 and Suzuki, Masato and 鈴木, 昌人 and Aoyagi, Seiji and 青柳, 誠司 and Arai, Yasuhiko and 新井, 泰彦}, journal = {Science and technology reports of Kansai University = 関西大学理工学研究報告}, month = {Mar}, note = {Examining differences in motion between individual persons involves predicting the three-dimensional (3D) future pose using an RGB-D camera. The 3D pose of a person is estimated based on depth data corresponding to the two-dimensional pose on the RGB image. Utilizing a neural network model, a sequence of 3D poses representing human motion is learned, and the future 3D pose is predicted from the sequential input of these 3D poses. The experiment evaluates the error in predicting the future 3D pose for individual persons. The evaluation involves basic motions like ‘standing’ and ‘sitting,’ which do not involve a change of location in everyday environments. The results highlight individual differences in motion and demonstrate the effectiveness of using personal motion data for learning and predicting the motion of individuals., Kansai University Fund for the Promotion and Enhancement of Education and Research, 2020. “Real-world Service Innovation through AI robot challenges.”, 2020年度 教育研究高度化促進費「AIロボットチャレンジを通した実世界サービスイノベーション」}, pages = {31--40}, title = {EVALUATION OF MOTION DIFFERENCE BETWEEN INDIVIDUAL PERSONS}, volume = {66}, year = {2024} }