Retrieve images using complex text descriptions by Neural Divide-and-Conquer Reasoning with NDCR
Retrieve images using complex text descriptions by Neural Divide-and-Conquer Reasoning with NDCR
A Neural Divide-and-Conquer Reasoning Framework for Image Retrieval from Linguistically Complex Text
arXiv paper abstract https://arxiv.org/abs/2305.02265
arXiv PDF paper https://arxiv.org/pdf/2305.02265.pdf
Pretrained Vision-Language Models (VLMs) have remarkable performance in image retrieval from text. However ... performance drops ... with linguistically complex texts ...
... regard ... complex texts as compound proposition texts composed of multiple simple proposition sentences and propose an end-to-end Neural Divide-and-Conquer Reasoning framework, dubbed NDCR.
It contains three main components: 1) Divide: ... divides the compound proposition text into simple proposition sentences ...
2) Conquer: ... a pretrained ... interactor achieves the interaction between ... sentences and images,
3) Combine: ... reasoner combines the ... states to obtain the final solution ...
... Experimental results and analyses indicate NDCR significantly improves performance in the complex image-text reasoning problem ...
Please like and share this post if you enjoyed it using the buttons at the bottom!
Stay up to date. Subscribe to my posts https://morrislee1234.wixsite.com/website/contact
Web site with my other posts by category https://morrislee1234.wixsite.com/website
Comments