A Computer Vision Framework for Automatic Description of Indian Monuments
Pushkar Shukla, Beena Rautela, Ankush Mittal · 2017
Monument recognition and description has emerged as a promising area of research. For any given image of a monument a question arises that up to what extend can a computer model describe the monument from that image?The main objective of the paper is to propose a framework which is capable of identifying multiple attributes from a single image of a monument. Four different attributes i.e. the class of the monument, the style of the architecture, the time period in which the monument was constructed and the type of the monument are taken into consideration. The paper proposes a framework that relies on Deep Convolutional Neural Networks (DCNN) for describing the monument in terms of the aforementioned attributes. The experiments have been performed on a dataset comprising of 6102 images of 117 Indian monuments. The model was able to achieve an accuracy greater than 80% for all the different set of experimentations. The results clearly indicate the usefulness of the framework.