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 Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

We present a semi-supervised method for panoptic segmentation based onConsInstancy regularisation, a novel strategy for semi-supervised learning. Itleverages completely unlabelled data by enforcing consistency between predictedinstance representations and semantic segmentations during training in order toimprove the segmentation performance. To this end, we also propose new types ofinstance representations that can be predicted by one simple forward paththrough a fully convolutional network (FCN), delivering a convenient andsimple-to-train framework for panoptic segmentation. More specifically, wepropose the prediction of a three-dimensional instance orientation map asintermediate representation and two complementary distance transform maps asfinal representation, providing unique instance representations for a panopticsegmentation. We test our method on two challenging data sets of both, hardenedand fresh concrete, the latter being proposed by the authors in this paperdemonstrating the effectiveness of our approach, outperforming the resultsachieved by state-of-the-art methods for semi-supervised segmentation. Inparticular, we are able to show that by leveraging completely unlabeled data inour semi-supervised approach the achieved overall accuracy (OA) is increased byup to 5% compared to an entirely supervised training using only labeled data.Furthermore, we exceed the OA achieved by state-of-the-art semi-supervisedmethods by up to 1.5%.

ConsInstancy: Learning Instance Representations for Semi-Supervised Panoptic Segmentation of Concrete Aggregate Particles