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

Semi-supervised learning is a challenging problem which aims to construct amodel by learning from a limited number of labeled examples. Numerous methodshave been proposed to tackle this problem, with most focusing on utilizing thepredictions of unlabeled instances consistency alone to regularize networks.However, treating labeled and unlabeled data separately often leads to thediscarding of mass prior knowledge learned from the labeled examples, andfailure to mine the feature interaction between the labeled and unlabeled imagepairs. In this paper, we propose a novel method for semi-supervised semanticsegmentation named GuidedMix-Net, by leveraging labeled information to guidethe learning of unlabeled instances. Specifically, we first introduce a featurealignment objective between labeled and unlabeled data to capture potentiallysimilar image pairs and then generate mixed inputs from them. The proposedmutual information transfer (MITrans), based on the cluster assumption, isshown to be a powerful knowledge module for further progressive refiningfeatures of unlabeled data in the mixed data space. To take advantage of thelabeled examples and guide unlabeled data learning, we further propose a maskgeneration module to generate high-quality pseudo masks for the unlabeled data.Along with supervised learning for labeled data, the prediction of unlabeleddata is jointly learned with the generated pseudo masks from the mixed data.Extensive experiments on PASCAL VOC 2012, PASCAL-Context and Cityscapesdemonstrate the effectiveness of our GuidedMix-Net, which achieves competitivesegmentation accuracy and significantly improves the mIoU by +7$\%$ compared toprevious state-of-the-art approaches.

GuidedMix-Net: Learning to Improve Pseudo Masks Using Labeled Images as Reference