Analysing Word Importance for Image Annotation

TL;DR

This paper investigates the importance of individual keywords in image annotation by considering their frequency and correlation, proposing a method to compute importance scores for more accurate annotations.

Contribution

It introduces a novel approach to determine keyword importance in image annotation based on frequency and correlation, addressing the limitation of treating all words equally.

Findings

Improved accuracy in image annotation by considering keyword importance

Effective scoring method for keywords with similar frequency

Enhanced relevance of annotated keywords in image retrieval

Abstract

Image annotation provides several keywords automatically for a given image based on various tags to describe its contents which is useful in Image retrieval. Various researchers are working on text based and content based image annotations [7,9]. It is seen, in traditional Image annotation approaches, annotation words are treated equally without considering the importance of each word in real world. In context of this, in this work, images are annotated with keywords based on their frequency count and word correlation. Moreover this work proposes an approach to compute importance score of candidate keywords, having same frequency count.