AWS Iot Greengrass for video detection

Do DLR image classification or a TensorFlow Lite image detection component also work for live video detection? I found that they get image from a image directory, but is it possible to upload the image from a web application? And is there a way for the application to fetch the result form the component?