Pose Animator takes a 2D vector illustration and animates its containing curves in real-time based on the recognition result from PoseNet and FaceMesh (both are TensorFlow.js models), as an SVG illustration. Using the camera feed or static images, it borrows the idea of skeleton-based animation from computer graphics, and applies it to vector characters.
It currently supports single-pose, single-face detection, and has been tested on Destkop Chrome & iOS Safari. (PoseNet model config - MobileNet V1, output stride 16, quant bytes 2).
Try it out now here