Deep Compositional Captioner

Please enable javascript in your browser.

Jahangiri, Saman | 2018

1674 Viewed

Type of Document: M.Sc. Thesis
Language: Farsi
Document No: 51126 (02)
University: Sharif University of Technology
Department: Mathematical Sciences
Advisor(s): Esfahani Zadeh, Mostafa; Kamali Tabrizi, Mostafa; Moghadasi, Jamshid
Abstract:
One of the most important applications of artificial intelligence, and especially deep learning is image captioning. Given an image, the task is to automatically produce a sentence, describing the image. Image captioning has several real world applications like helping the blind understanding the images, generating automatic captions for the social media, etc. In the past, several different methods for image captioning have been used, but after the emergence of deep learning, like many other areas, image captioning algorithms have been improved significantly. In this thesis, I talk about a specific method for image captioning, called ”Deep Compositional Captioning. In this method, at first image features are being extracted from the given image, by a deep classification convolutional neural network. Then, another deep neural networkthe language model- which has been trained on text corpus, uses these features to produce a sentence about the image. At last, I will talk about a method for describing the ”novel” objects for which there is no captioning in the training data
Keywords:
Deep Learning ; Images Classification ; Natural Language Processing ; Object Detection ; Image Captioning ; Word Embedding