@techreport{oai:ipsj.ixsq.nii.ac.jp:00211571, author = {尾関, 剛成 and 青木, 直史 and 土橋, 宜典 and 池田, 研一 and 安田, 寛 and Kosei, Ozeki and Naofumi, Aoki and Yoshinori, Dobayashi and Kenichi, Ikeda and Hiroshi, Yasuda}, issue = {62}, month = {Jun}, note = {音波通信は,電波通信と比べると,環境雑音や反響音,残響音,ドップラー効果などの外乱により信号音が歪みやすく,情報識別に誤りが生じやすいことが問題になっている.本研究では,音波通信のさらなる高度化を目指し,深層学習を利用して外乱環境下から信号を検討する識別器を設計し,その性能評価を行った結果について述べる., The problem with sound wave communication is that it causes more incorrect identification than radio wave communication. It is caused by the distortion of sound signals due to disturbances such as environmental noise, reverberation, reverberation, and Doppler effect. In this study, aiming at further sophistication of sound wave communication, we design a classifier that discriminates signals from a disturbance environment using deep learning, and describe the results of performance evaluation.}, title = {深層学習による超音波信号分類システムの開発}, year = {2021} }