What we are doing here is to use Twitter API and Google API to do image and vedio label recognition
Use twitter API to catch image from an account of twitter.
Because the picture we get may not be well formed, so we need to fit those images into same size before we make it into an vedio
Use google API to do the image and vedio recognition and then get the description of the image and vedio