Computer Science > Computer Vision and Pattern Recognition
[Submitted on 31 Jan 2022 (v1), revised 23 Jun 2023 (this version, v3), latest version 22 Aug 2023 (v5)]
Title:Deep Learning Approaches on Image Captioning: A Review
View PDFAbstract:Image captioning is a challenging research area that aims to generate natural language descriptions for visual content. The advent of deep learning and more recently vision-language pre-training techniques has revolutionized the field, leading to more sophisticated methods and improved performance. This survey paper provides a structured review of deep learning methods in image captioning by providing a comprehensive taxonomy and discussing each method category in detail. We also discuss the widely-used datasets and evaluation metrics created to assess the performance of image captioning models. We highlight the challenges faced in the field, such as the object hallucination problem, missing context, illumination conditions, contextual understanding, and referring expressions. We rank various deep learning methods in terms of their performance according to established evaluation metrics. In addition to identifying the current state of the art, we suggest potential future directions for research in this area, such as mitigating the information misalignment problem between the image and text modalities, overcoming the dataset bias, incorporating vision-language pre-training methods for caption generation, and developing better evaluation tools to measure the quality of image captions.
Submission history
From: Taraneh Ghandi [view email][v1] Mon, 31 Jan 2022 00:39:37 UTC (4,144 KB)
[v2] Thu, 17 Nov 2022 16:31:29 UTC (3,078 KB)
[v3] Fri, 23 Jun 2023 23:24:18 UTC (2,562 KB)
[v4] Sat, 12 Aug 2023 16:59:44 UTC (2,565 KB)
[v5] Tue, 22 Aug 2023 17:50:41 UTC (2,564 KB)
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
Connected Papers (What is Connected Papers?)
CORE Recommender (What is CORE?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.