Recurrent Attentional Networks for Saliency Detection

TL;DR

This paper introduces RACDNN, a recurrent attentional network that improves saliency detection by iteratively focusing on image regions and learning context-aware features, outperforming existing methods.

Contribution

The novel RACDNN model combines spatial transformers and recurrent units to address scale variation and context modeling in saliency detection.

Findings

RACDNN outperforms state-of-the-art methods on multiple datasets.

It effectively handles objects of multiple scales.

The model demonstrates improved saliency refinement through iterative attention.

Abstract

Convolutional-deconvolution networks can be adopted to perform end-to-end saliency detection. But, they do not work well with objects of multiple scales. To overcome such a limitation, in this work, we propose a recurrent attentional convolutional-deconvolution network (RACDNN). Using spatial transformer and recurrent network units, RACDNN is able to iteratively attend to selected image sub-regions to perform saliency refinement progressively. Besides tackling the scale problem, RACDNN can also learn context-aware features from past iterations to enhance saliency refinement in future iterations. Experiments on several challenging saliency detection datasets validate the effectiveness of RACDNN, and show that RACDNN outperforms state-of-the-art saliency detection methods.