Child Emotion Recognition (Crying/Tantr…

Using fixed cameras (and optional microphones) at home or in kindergartens, AI multimodal analysis recognizes a child's facial expressions (eyebrow/eye shape, mouth-corner curvature), cry-sound features (pitch, frequency, duration), and body-motion amplitude (waving, stomping, curling up) in real time, and jointly identifies the child's typical emotional state: happy, sad, angry, fearful, etc. | 通过家庭或幼儿园内的固定摄像头（及可选麦克风），利用AI多模态分析技术实时分析儿童的面部表情（如眉眼形态、嘴角弧度）、哭声音频特征（音调、频率、持续时间）以及肢体动作幅度（挥手、跺脚、蜷缩等），综合识别出儿童当前的典型情绪状态：快乐、悲伤、愤怒、恐惧等。