Child Drowsiness / Fatigue Detection | 儿童打瞌睡/疲劳检测

Using a fixed camera in the classroom or above the home desk, the system analyzes the child's (student's) facial video in real time, detecting eye closure ratio (PERCLOS — the proportion of time eyes are closed more than 80% within a unit time), head-nodding frequency (rapid downward nod followed by lift), and changes in eye-region glossiness, and computes a comprehensive fatigue index (0-100). The skill helps teachers or parents detect drowsiness in time and adjust learning schedules or remind the child to rest. Application scenarios: classrooms, home desks, online classes. The system monitors in real time and, when the fatigue index exceeds a threshold, pushes reminders or triggers voice prompts (e.g., 'kid, time to take a break'). Skill features: safeguarding sleep health. | 通过教室或家庭书桌的固定摄像头，实时分析儿童（学生）的面部视频，检测眼部闭合比例（PERCLOS，单位时间内眼睛闭合超过80%的时间占比）、点头动作频率（头部快速下点后抬起）以及眼部区域的光泽度变化，综合计算疲劳指数（0-100）。该技能可帮助教师或家长及时发现儿童困倦状态，调整学习安排或提醒休息。应用场景：教室、家庭书桌、在线课堂。系统实时监测，当疲劳指数超过阈值时，推送提醒或触发语音提示（如'小朋友，休息一下吧'）。技能特点：保障睡眠健康。

by @smyx-sunjinhui·MIT-0