Video Analytics-based Ingredient Detection for Smart Chef using YOLOv3 and ImageAI

With the boom in internet usage every aspect of our daily lives has experienced a drastic change over a period of time. Searching the internet for food recipe videos is one of the most common searches around the world. Search results from internet are shown based on the keywords the video is tagged with and often it is observed that the videos are not much relevant and does not grab user's immediate attention. Our work is concerned with analyzing the objects and contents of videos to find the actual ingredients involved in them. The objective is to provide a list of the objects involved in the video. In this work, we have used YOLOv3 and ImageAI to detect objects from the video files. We consider using the COCO dataset for the analysis. So, the project is concerned about identifying relevant recipe videos conveniently using video analysis.