We propose cross-media multi-level alignment to explore global, local and relation alignments across different media types, which can mutually boost to learn ...

Multi-level alignment is fully exploited with relation attention network between the global original instances, local fine-grained patches as well as their ...

Apr 25, 2018 · Second, we propose cross-media multi-level alignment to explore global, local and relation alignments across different media types, which can ...

We aim to not only exploit cross-media fine-grained local information, but also capture the intrinsic relation information, which can provide complementary ...

We aim to not only exploit cross-media fine-grained local information, but also capture the intrinsic relation information, which can provide complementary ...

This work proposes Cross-media Relation Attention Network (CRAN) with multi-level alignment to explore global, local and relation alignments across ...

Qi et al. [32] developed a cross-media relation attention network with three branches to find global, local, and relation alignment (multilevel alignment) ...

The key of image and sentence matching is to accurate- ly measure the visual-semantic similarity between an image and a sentence.

This paper presents a novel Cross-modal Feature Alignment based Hybrid Attentional Generative Adversarial Networks (CFA-HAGAN) for text-to-image synthesis.

Apr 25, 2018 · Multi-level alignment is fully exploited with relation attention network between the global original instances, local fine-grained patches as ...