Performance Comparison of AI Models for Digital Image captioning

Abstract

Digital Image captioning field deals with the concept of interpreting a given image by the application of different Artificial Intelligence (AI) based algorithms. These kind of systems always considered the input in the form of image and replying with an output representing some text or speech samples. This output explains the content present in the image. These kind of automated systems proven good in the areas like medical image processing, online review systems, speech conversion for visually impaired persons etc. This paper represents a comparative study among four different AI models: Decision Tree (DT), Random Forest (RF), Support Vector Machine (SVM) and kNN (k-Nearest Neighbor) in context of image captioning activity. Random scenes in the form of input images has been consider to show the part of diverse nature of these algorithms. It has been observed that SVM classifier outshine as compared to rest of the models with the overall accuracy of 98%.