Extracting general concepts, rather than specific phrases, from documents and contracts is challenging. It's even more complicated when applied to scanned documents containing handwritten annotations. We describe using object detection and OCR with Azure ML Package for Computer Vision and Cognitive Services API.
