Optionally, MediaPipe Pose can predicts a full-body segmentation mask represented as a two-class segmentation (human or background).
確定! 回上一頁