Evaluations of training paradigms in neural image captioning

This project presents an implementation of 2 neural image captioning model which shall achieve results comparable to state of the art models. The 2 models are composed of different training paradigm, cross entropy training or self-critical training: • Model-1: Bottom-up attention with Self-Critical...

Full description

Saved in:
Bibliographic Details
Main Author: Lee, Si Min
Other Authors: Zhang Hanwang
Format: Final Year Project
Language:English
Published: Nanyang Technological University 2019
Subjects:
Online Access:https://hdl.handle.net/10356/136508
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
id sg-ntu-dr.10356-136508
record_format dspace
spelling sg-ntu-dr.10356-1365082019-12-20T07:44:00Z Evaluations of training paradigms in neural image captioning Lee, Si Min Zhang Hanwang School of Computer Science and Engineering hanwangzhang@ntu.edu.sg Engineering Engineering::Computer science and engineering This project presents an implementation of 2 neural image captioning model which shall achieve results comparable to state of the art models. The 2 models are composed of different training paradigm, cross entropy training or self-critical training: • Model-1: Bottom-up attention with Self-Critical Training • Model-2: Bottom-up attention with cross entropy Training There will be a comparison and evaluation of the results against each model and related neural image captioning models to determine the best performing model. Bachelor of Engineering (Computer Science) 2019-12-20T07:44:00Z 2019-12-20T07:44:00Z 2019 Final Year Project (FYP) https://hdl.handle.net/10356/136508 en application/pdf Nanyang Technological University
institution Nanyang Technological University
building NTU Library
country Singapore
collection DR-NTU
language English
topic Engineering
Engineering::Computer science and engineering
spellingShingle Engineering
Engineering::Computer science and engineering
Lee, Si Min
Evaluations of training paradigms in neural image captioning
description This project presents an implementation of 2 neural image captioning model which shall achieve results comparable to state of the art models. The 2 models are composed of different training paradigm, cross entropy training or self-critical training: • Model-1: Bottom-up attention with Self-Critical Training • Model-2: Bottom-up attention with cross entropy Training There will be a comparison and evaluation of the results against each model and related neural image captioning models to determine the best performing model.
author2 Zhang Hanwang
author_facet Zhang Hanwang
Lee, Si Min
format Final Year Project
author Lee, Si Min
author_sort Lee, Si Min
title Evaluations of training paradigms in neural image captioning
title_short Evaluations of training paradigms in neural image captioning
title_full Evaluations of training paradigms in neural image captioning
title_fullStr Evaluations of training paradigms in neural image captioning
title_full_unstemmed Evaluations of training paradigms in neural image captioning
title_sort evaluations of training paradigms in neural image captioning
publisher Nanyang Technological University
publishDate 2019
url https://hdl.handle.net/10356/136508
_version_ 1681036893058236416