Image Captioning Bot Using Deep Learning

Image Captioning Bot Using Deep Learning

Rohan Kumar 1 Shimona Wadhwa 2  Raghav Bansal 3

---------------------------------------------------------------------***---------------------------------------------------------------------

Abstract - In this work, we concentrate on image captioning, one type of visual recognition. The model's goal is to create a caption for a picture. Its many uses have garnered a lot of attention within the past ten years. We are utilising deep learning techniques in our project to automatically create image captions. CNN is the encoder and RNN is the decoder for the Flick8k dataset. The ResNet model is what we are utilising to extract features. For those who are blind or visually challenged, this approach is quite helpful in creating captions for posts on social media, among other things.

Key Words: Deep Learning, Flickr8k, CNN, RNN, ResNet50, Encoder-Decoder Model