Generating Image Descriptions using Attention Mechanism
Ms. Malge Shraddha · International Journal for Research in Applied Science and Engineering Technology · 2021
In simple terms, Image Captioning means the task of describing an image. Description comprises all the details which describe the objects within the image and their spatial connectivity.This task is performed very effectively by a human being and there are various methods to make a machine do so.In this project, we develop the encoder -decoder model to caption the image, to improve its accuracy we are using the visual attention mechanism.We used the Local/Bahdanau attention mechanism which attends to only a subset of words and is computationally simpler than global attention.We are using BLEU as a metric to evaluate the correctness of the generated caption.