• artificial intelligence
• computer vision
• deep learning
• document layout analysis
• faster r-cnn
• google cloud platform
• information extraction
• neural network
• object detection
• region proposal
• theater scripts
• transfer learning
• yolo

The computer vision and object detection techniques developed in recent years are dominating the state of the art, and are increasingly applied to document layout analysis resolutions. This paper wants to offer a method to process digitized documents for the purpose of extracting meaningful information. By fine-tuning object detectors such as Faster R-CNN and YOLO, we attempt to identify text sections of interest with bounding boxes and classify them into a specific category depending on the context in which the document is placed. The deep learning model is implemented using the Python programming language, and is eventually integrated into the back-end of a web application hosted on the Google Cloud Platform infrastructure.