appsilonoffice - Tumblr blog

[appsilonoffice]

What is YOLO object detection?

Would you like to know what is YOLO object detection? Then read this post! Lets's start with image classification, as you may already know it's an example of convolution neutral network. 

 YOLO - meaning

It's an effective real-time object recognition algorithm, it was first described back in the year 2015. The acronym YOLO stands for: You Only Look Once. One of the open-source implementations of this algorithm is Darknet.

Object detection - how does it work?

First it starts with image classification, for example cat, car, human etc. one picture can belong to only one category. Then comes object localization, the algorithm is looking for an answer to the question: where the object from the picture is. The third is object detection, the algorithm is finding all the object presented in an image and then is drawing the co-called - bounding boxes. In some situation, bounding boxes aren't the most precise, and we need to know the exact boundaries of the object. Determination of these boundaries is called instance segmentation, but this is a topic for some other time. If you would like to know more about what is YOLO object detection, just click on the link to the website!