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

Generalizing models trained on normal visual conditions to target domainsunder adverse conditions is demanding in the practical systems. One prevalentsolution is to bridge the domain gap between clear- and adverse-conditionimages to make satisfactory prediction on the target. However, previous methodsoften reckon on additional reference images of the same scenes taken fromnormal conditions, which are quite tough to collect in reality. Furthermore,most of them mainly focus on individual adverse condition such as nighttime orfoggy, weakening the model versatility when encountering other adverseweathers. To overcome the above limitations, we propose a novel framework,Visibility Boosting and Logit-Constraint learning (VBLC), tailored for superiornormal-to-adverse adaptation. VBLC explores the potential of getting rid ofreference images and resolving the mixture of adverse conditionssimultaneously. In detail, we first propose the visibility boost module todynamically improve target images via certain priors in the image level. Then,we figure out the overconfident drawback in the conventional cross-entropy lossfor self-training method and devise the logit-constraint learning, whichenforces a constraint on logit outputs during training to mitigate this painpoint. To the best of our knowledge, this is a new perspective for tacklingsuch a challenging task. Extensive experiments on two normal-to-adverse domainadaptation benchmarks, i.e., Cityscapes -> ACDC and Cityscapes ->FoggyCityscapes + RainCityscapes, verify the effectiveness of VBLC, where itestablishes the new state of the art. Code is available athttps://github.com/BIT-DA/VBLC.

VBLC: Visibility Boosting and Logit-Constraint Learning for Domain Adaptive Semantic Segmentation under Adverse Conditions