Image Caption Generation Using Recurrent Convolutional Neural Network

This paper presents a residual learning (RL) approach to generate automated captions for any given image. In this approach, a convolutional neural network (CNN) is employed to extract the spectral and spatial characteristics of the image, which is essential to solve the caption generation problem, which necessitates the use of CNN. In addition to this, we consider the nuanced quality of language by incorporating an image annotation generator into the system that has been recommended. The results of the experiments that have been presented here provide convincing evidence that the developed model is an improvement upon the various approaches to image captioning that are currently being used. © 2024, Ismail Saritas. All rights reserved.