Semantic Similarity based measurement for Lung’s infection imagery using Deep Learning

Image recognition is one of the world’s fastest growing technologies is Image recognition. Its use in security, monitoring systems has made it the matter of talk in the recent times. There are plentiful of imaging applications which require image similarity as a part of their procedure. Though the applications are quite different, but the process is same. To discover the steps which are crucial in recognizing similar images is one main challenge in image recognition. Preprocessing and feature extraction have been given the importance by past studies. However, the role of similarity measure is still not clear. This paper provides the results which indicates that for recognizing similar images the similarity measures are important. This paper proposes one of the popular similarity metrics called cosine similarity.