First impression assessment of digital human applicant images generated with posture prompts and text prompts

[Abstract]

We investigated whether a digital human applicant generated by a combination of posture prompts and additional text prompts elicited a good or bad first impression during an interview in a virtual space. Existing analytical studies have not evaluated whether a combination of posture prompts and additional text prompts can improve an interviewer's first impression of a digital human applicant in an interview setting. To examine this issue, we generated images of digital human applicants by combining the presence or absence of posture prompts and the presence or absence of additional text prompts. We conducted subjective assessments in which participants simulating interviewers in a virtual space reported their first impressions of the images. The experimental results demonstrated that interviewers' first impressions of a digital human applicant generated by combining posture prompts and additional text prompts were better than their first impressions of a digital human applicant generated without prompts.

[Publications]

• Wataru Ganaha, Shota Hioki, Masashi Nishiyama,
  First impression assessment of digital human applicant images generated with posture prompts and text prompts, Proceedings of 27th International Conference on Human-Computer Interaction (HCII), vol. DHM, pp. 185 - 199, June 2025. [Slides]
• 我那覇 航, 日置 翔太, 西山 正志,
  姿勢プロンプトとテキストプロンプトとの組み合わせにより画像生成された面接志願者の第一印象の主観評価,
  HCGシンポジウム, A-1-3, pp. 1 - 8, December 2024. [Slides]