A comparative analysis of face annotation schemes for web facial images in social networks

This paper aimed to compare the analysis of two different face annotation techniques. There is a sudden powerful research in mining weakly labeled web facial images on the internet to tackle the existing longtime research challenges in computer vision and image processing. In this paper, unified learning framework for auto face annotation by mining web facial images and another scheme Mining weakly labeled web facial images for search-based face annotation are analyzed and the former designed with two learning schemes such as inductive and transductive learning´s which are used for this framework. For inductive, weak label laplacian support vector machine is proposed and for transductive weak label regularized local coordinate coding is used. By combining inductive and transductive learning schemes produces a maximum efficacy. For search -based face annotation two algorithms such as ULR algorithm is used to refine the weak labels of web facial images using machine learning techniques and clustering algorithm is used to improve the scalability of the web facial images.