The demos provide an inference pipeline for three multi-channel scenarios: face detection, human pose estimation and object detection yolov3. For more information, refer to the corresponding pages: