Abstract: This paper describes a machine learning approach for visual object detection which is capable of processing images extremely rapidly and achieving high detection rates. This work is distinguished by two key contributions. The first is the Convolutional Neural Networks (CNN) learning algorithm that will help us with reducing computational time to a very large degree that makes the process extremely fast. The second contribution is training data and testing data. As we know from machine learning definition (the ability of a machine to learn without being explicitly programmed), a machine can learn and grow over time. With enough training data, we will then train machine to be able to predict objects with higher level of accuracy. The next step would be to test it. The more the test data the better. With the help of these two contributions, we will get a machine that is much more accurate and way faster than many existing machines.
Keywords: Object Detection, Machine Learning, CNN algorithm, training data and testing data
Copyright Form