情動喚起動画視聴直後に表出した作為的な笑顔における情動判別のための顔特徴点解析

書誌事項

タイトル別名
  • Face Landmark Analysis for Emotion Discrimination on Fake Smile Expressing just after Watching Emotion-eliciting Film

抄録

<p>The purpose of this study is to clarify feature for discriminating emotion while expressing fake smile. First, we experimented to record face expression in video. The flow of the experiment was following; we first displayed one of emotion-eliciting films, and participant expressed (fake) smile just after watching the film. At the end of the experiment, we let them assess what emotion they felt with two self-report measures. We prepared six emotion-eliciting films in order to record various face expression with various emotion. Then, we obtained face landmark positions from the face on the recorded video using a method proposed by Kazemi et al(1). We focused that eyes and mouth corners move well while smiling, and defined 15 feature values related to that landmarks. Furthermore, we grouped the feature values in emotion scores by clustering. There are significant differences in eye area between happiness and disgust, and in some mouth corners’ feature values between happiness and fear as a result of comparison of clusters. we considered these differences resulted from disgust/fear, but negative emotion.</p>

収録刊行物

  • 年次大会

    年次大会 2019 (0), S12103P-, 2019

    一般社団法人 日本機械学会

詳細情報 詳細情報について

問題の指摘

ページトップへ