Evaluation on Comfortable Arousal in Autonomous Driving Using Physiological Indexes

Naoki Sakashita, Narumon Jadram, Peeraya Sripian, Tipporn Laohakangvalvit, Midori Sugaya

研究成果: Conference contribution

抄録

At level 3 of autonomous driving, the driver has to take over driving when the system requires. In automatic driving, the arousal level tends to decrease. Drowsiness or less arousal is the leading cause of car accidents. For safety, it is necessary to increase the arousal level before driving. Moreover, due to the emotional state effect on the driving performance, it is important to consider comfort while improving the driver's arousal level. Previous studies proposed the comfortable arousal model based on physiological signals to evaluate arousal and comfort during autonomous driving. However, the accuracy evaluation using this model has not been sufficiently performed. This study aims to construct a more accurate and reliable comfortable arousal model. We explore various physiological indexes and calculate feature importance using the random forest method to achieve our goal. Then we compare and validate the evaluation accuracy with the subjective evaluation score against the previous comfortable model proposed. The result shows that the proposed method has more accurate than the methods of the previous method. However, the improved accuracy is still not very high, so we need to consider creating a comfortable arousal model.

本文言語English
ホスト出版物のタイトルArtificial Intelligence in HCI - 3rd International Conference, AI-HCI 2022, Held as Part of the 24th HCI International Conference, HCII 2022, Proceedings
編集者Helmut Degen, Stavroula Ntoa
出版社Springer Science and Business Media Deutschland GmbH
ページ305-316
ページ数12
ISBN(印刷版)9783031056420
DOI
出版ステータスPublished - 2022
イベント3rd International Conference on Artificial Intelligence in HCI, AI-HCI 2022 Held as Part of the 24th HCI International Conference, HCII 2022 - Virtual, Online
継続期間: 2022 6月 262022 7月 1

出版物シリーズ

名前Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
13336 LNAI
ISSN(印刷版)0302-9743
ISSN(電子版)1611-3349

Conference

Conference3rd International Conference on Artificial Intelligence in HCI, AI-HCI 2022 Held as Part of the 24th HCI International Conference, HCII 2022
CityVirtual, Online
Period22/6/2622/7/1

ASJC Scopus subject areas

  • 理論的コンピュータサイエンス
  • コンピュータサイエンス一般

フィンガープリント

「Evaluation on Comfortable Arousal in Autonomous Driving Using Physiological Indexes」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。

引用スタイル