We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.CV

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Computer Vision and Pattern Recognition

Title: Focus! Relevant and Sufficient Context Selection for News Image Captioning

Abstract: News Image Captioning requires describing an image by leveraging additional context from a news article. Previous works only coarsely leverage the article to extract the necessary context, which makes it challenging for models to identify relevant events and named entities. In our paper, we first demonstrate that by combining more fine-grained context that captures the key named entities (obtained via an oracle) and the global context that summarizes the news, we can dramatically improve the model's ability to generate accurate news captions. This begs the question, how to automatically extract such key entities from an image? We propose to use the pre-trained vision and language retrieval model CLIP to localize the visually grounded entities in the news article and then capture the non-visual entities via an open relation extraction model. Our experiments demonstrate that by simply selecting a better context from the article, we can significantly improve the performance of existing models and achieve new state-of-the-art performance on multiple benchmarks.
Comments: Findings of EMNLP 2022
Subjects: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL)
Cite as: arXiv:2212.00843 [cs.CV]
  (or arXiv:2212.00843v1 [cs.CV] for this version)

Submission history

From: Mingyang Zhou [view email]
[v1] Thu, 1 Dec 2022 20:00:27 GMT (2074kb,D)

Link back to: arXiv, form interface, contact.