English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

We hypothesize that end-to-end neural image captioning systems work seeminglywell because they exploit and learn `distributional similarity' in a multimodalfeature space by mapping a test image to similar training images in this spaceand generating a caption from the same space. To validate our hypothesis, wefocus on the `image' side of image captioning, and vary the input imagerepresentation but keep the RNN text generation component of a CNN-RNN modelconstant. Our analysis indicates that image captioning models (i) are capableof separating structure from noisy input representations; (ii) suffer virtuallyno significant performance loss when a high dimensional representation iscompressed to a lower dimensional space; (iii) cluster images with similarvisual and linguistic information together. Our findings indicate that ourdistributional similarity hypothesis holds. We conclude that regardless of theimage representation used image captioning systems seem to match images andgenerate captions in a learned joint image-text semantic subspace.

End-to-end Image Captioning Exploits Multimodal Distributional Similarity