Image Captioning is the process of creating a text description of an image. It uses both Natural Language Processing (NLP) and Computer Vision to generate the captions. The image captioning task is done by combining the detection process when the descriptions consist of a single word like cat, skateboard, etc. and Image Captioning when one predicted region covers the full image, for example cat riding a skateboard. To address the localization and description task together, we propose a Fully Convolution Localization Network that processes a picture with a single forward pass which can be consistently trained in a single round of optimization. To process an image, first the input image is processed using CNN. Then Localization Layer proposes regions and includes a region detection network adopted from faster R-CNN and captioning network. The model directly combines the faster R-CNN framework for region detection and long short-term memory (LSTM) for captioning

Keywords : Image Caption, Recurrent Neural Network, Long short-term memory, Convolution Neural Network, Faster R-CNN, Natural Language Processing