Scholarlink

<p>Humans navigate crowded environments by anticipating others’ movements. When such mutual anticipation fails, this may result in collisions or near-collisions. Previous research on mutual anticipation in collision avoidance has highlighted the importance of kinematic information. However, it remains unclear how full-body kinematics support mutual anticipation in truly interactive near-collision avoidance scenarios.　We introduce cross-recurrence quantification analysis (CRQA) as a novel method to quantify the degree of mutual hesitation (i.e., spatiotemporal overlap between walkers) in near-collision avoidance. Additionally, we examined which pre-avoidance, inter-participant kinematic relationships predict the degree of mutual hesitation. Forty-five dyads (i.e., 90 participants) walked toward each other and aimed at avoiding collision in two differently sized avoidance zones (20&#xa0;cm, 80&#xa0;cm). Results revealed that the mean Recurrence Rate in mutual hesitation trials was 37.97 (± 23.2) and 41.95 (± 25.5) in the respective collision avoidance zones. Statistical Parametric Mapping showed that inter-participant head-angle differences diverged earlier than the shoulder or pelvis in non-mutual hesitation trials. Smaller head-angle differences 0.3–0.2s (20&#xa0;cm) and 0.2–0.1s (80&#xa0;cm) before avoidance onset predicted a higher degree of mutual hesitation. We conclude that failures in inferring avoidance direction from head kinematics may be linked to mutual anticipation errors that result in mutual hesitation.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Quantifying mutual hesitation and identifying kinematic predictors in near-collision avoidance in walking

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月7日

查看全文
  • Kazuyuki Sato,
  • Johannes Keck,
  • Rouwen Cañal-Bruland

摘要

Humans navigate crowded environments by anticipating others’ movements. When such mutual anticipation fails, this may result in collisions or near-collisions. Previous research on mutual anticipation in collision avoidance has highlighted the importance of kinematic information. However, it remains unclear how full-body kinematics support mutual anticipation in truly interactive near-collision avoidance scenarios.　We introduce cross-recurrence quantification analysis (CRQA) as a novel method to quantify the degree of mutual hesitation (i.e., spatiotemporal overlap between walkers) in near-collision avoidance. Additionally, we examined which pre-avoidance, inter-participant kinematic relationships predict the degree of mutual hesitation. Forty-five dyads (i.e., 90 participants) walked toward each other and aimed at avoiding collision in two differently sized avoidance zones (20 cm, 80 cm). Results revealed that the mean Recurrence Rate in mutual hesitation trials was 37.97 (± 23.2) and 41.95 (± 25.5) in the respective collision avoidance zones. Statistical Parametric Mapping showed that inter-participant head-angle differences diverged earlier than the shoulder or pelvis in non-mutual hesitation trials. Smaller head-angle differences 0.3–0.2s (20 cm) and 0.2–0.1s (80 cm) before avoidance onset predicted a higher degree of mutual hesitation. We conclude that failures in inferring avoidance direction from head kinematics may be linked to mutual anticipation errors that result in mutual hesitation.