I found this tutorial from google - creating-object-detection-application-tensorflow. After looking into its github page --> object_detection_app --> app.py we only need to run detect_objects(image_path) function every single time we want to detect an object.