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

Objects at different spatial positions in an image exhibit different scales.Adaptive receptive fields are expected to capture suitable ranges of contextfor accurate pixel level semantic prediction. Recently, atrous convolution withdifferent dilation rates has been used to generate features of multi-scalesthrough several branches which are then fused for prediction. However, there isa lack of explicit interaction among the branches of different scales toadaptively make full use of the contexts. In this paper, we propose aContent-Adaptive Scale Interaction Network (CASINet) to exploit the multi-scalefeatures for scene parsing. We build CASINet based on the classic AtrousSpatial Pyramid Pooling (ASPP) module, followed by a proposed contextual scaleinteraction (CSI) module, and a scale adaptation (SA) module. Specifically, inthe CSI module, for each spatial position of some scale, instead of beinglimited by a fixed set of convolutional filters that are shared acrossdifferent spatial positions for feature learning, we promote the adaptivity ofthe convolutional filters to spatial positions. We achieve this by the contextinteraction among the features of different scales. The SA module explicitlyand softly selects the suitable scale for each spatial position and eachchannel. Ablation studies demonstrate the effectiveness of the proposedmodules. We achieve state-of-the-art performance on three scene parsingbenchmarks Cityscapes, ADE20K and LIP.

CaseNet: Content-Adaptive Scale Interaction Networks for Scene Parsing