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

This paper exploits the intrinsic features of urban-scene images and proposesa general add-on module, called height-driven attention networks (HANet), forimproving semantic segmentation for urban-scene images. It emphasizesinformative features or classes selectively according to the vertical positionof a pixel. The pixel-wise class distributions are significantly different fromeach other among horizontally segmented sections in the urban-scene images.Likewise, urban-scene images have their own distinct characteristics, but mostsemantic segmentation networks do not reflect such unique attributes in thearchitecture. The proposed network architecture incorporates the capabilityexploiting the attributes to handle the urban scene dataset effectively. Wevalidate the consistent performance (mIoU) increase of various semanticsegmentation models on two datasets when HANet is adopted. This extensivequantitative analysis demonstrates that adding our module to existing models iseasy and cost-effective. Our method achieves a new state-of-the-art performanceon the Cityscapes benchmark with a large margin among ResNet-101 basedsegmentation models. Also, we show that the proposed model is coherent with thefacts observed in the urban scene by visualizing and interpreting the attentionmap. Our code and trained models are publicly available athttps://github.com/shachoi/HANet

Cars Can't Fly up in the Sky: Improving Urban-Scene Segmentation via Height-driven Attention Networks