Elm: From Documents to Traffic Enforcement: Applying object detection framework with YOLOV5/PyTorch
You Only Look Once (YOLO) is one of the most frequently used and popular algorithms for AI scientists and engineers working on various object detection problems. This demo explains a complete, end-to-end dockerized object detection pipeline presenting the data extraction, annotation, training, modeling, and testing stages for two real-world computer vision problems. The first part of the demo presents a Flask-based project detailing the development of a landmark/object detection algorithm for document verification. The second part then presents a different use-case covering the traffic enforcement domain where a double-parking detection application is explained and demonstrated.