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

Most prior semantic segmentation methods have been developed for day-timescenes, while typically underperforming in night-time scenes due toinsufficient and complicated lighting conditions. In this work, we tackle thischallenge by proposing a novel night-time semantic segmentation paradigm, i.e.,disentangle then parse (DTP). DTP explicitly disentangles night-time imagesinto light-invariant reflectance and light-specific illumination components andthen recognizes semantics based on their adaptive fusion. Concretely, theproposed DTP comprises two key components: 1) Instead of processinglighting-entangled features as in prior works, our Semantic-OrientedDisentanglement (SOD) framework enables the extraction of reflectance componentwithout being impeded by lighting, allowing the network to consistentlyrecognize the semantics under cover of varying and complicated lightingconditions. 2) Based on the observation that the illumination component canserve as a cue for some semantically confused regions, we further introduce anIllumination-Aware Parser (IAParser) to explicitly learn the correlationbetween semantics and lighting, and aggregate the illumination features toyield more precise predictions. Extensive experiments on the night-timesegmentation task with various settings demonstrate that DTP significantlyoutperforms state-of-the-art methods. Furthermore, with negligible additionalparameters, DTP can be directly used to benefit existing day-time methods fornight-time segmentation.

Disentangle then Parse:Night-time Semantic Segmentation with Illumination Disentanglement