Evaluation of Content Using Machine Learning

Machine learning is one of the most widely used approaches in current day for effective analysis and prediction purpose in the domain of image processing applications. This paper thus makes a study of existing image retrieval methods. It has been a challenge to implement human recognition method in machines for a long time now. In this process number of methods and procedures are developed to implement the human recognition system in machines. It is however still not possible to achieve the accuracy human are bestowed with. There are number of ways in which different researchers have attempted this task. The paper describes the method along with the parameters used, processing method, and the accuracy achieved. The latest method being machine learning, this study concentrates on the development of supervised learning and thereof the progress made by the unsupervised learning methods. The two techniques Text-based Image Retrieval and Content-Based Image Retrieval are the important retrieval methods dealt with.