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 Tools annual

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 monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

In recent years, the development of instance segmentation has garneredsignificant attention in a wide range of applications. However, the training ofa fully-supervised instance segmentation model requires costly bothinstance-level and pixel-level annotations. In contrast, weakly-supervisedinstance segmentation methods (i.e., with image-level class labels or pointlabels) struggle to satisfy the accuracy and recall requirements of practicalscenarios. In this paper, we propose a novel paradigm called synthetic instancesegmentation (SISeg), which achieves Instance Segmentation results from imagemasks predicted using off-the-shelf semantic segmentation models. SISeg doesnot require training a semantic or/and instance segmentation model and avoidsthe need for instance-level image annotations. Therefore, it is highlyefficient. Specifically, we first obtain a semantic segmentation mask of theinput image via a trained semantic segmentation model. Then, we calculate adisplacement field vector for each pixel based on the segmentation mask, whichcan indicate representations belonging to the same class but differentinstances, i.e., obtaining the instance-level object information. Finally,instance segmentation results are obtained after being refined by a learnablecategory-agnostic object boundary branch. Extensive experimental results on twochallenging datasets and representative semantic segmentation baselines(including CNNs and Transformers) demonstrate that SISeg can achievecompetitive results compared to the state-of-the-art fully-supervised instancesegmentation methods without the need for additional human resources orincreased computational costs. The code is available at: SISeg

Synthetic Instance Segmentation from Semantic Image Segmentation Masks