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

Current instance segmentation methods rely heavily on pixel-level annotatedimages. The huge cost to obtain such fully-annotated images restricts thedataset scale and limits the performance. In this paper, we formally addresssemi-supervised instance segmentation, where unlabeled images are employed toboost the performance. We construct a framework for semi-supervised instancesegmentation by assigning pixel-level pseudo labels. Under this framework, wepoint out that noisy boundaries associated with pseudo labels are double-edged.We propose to exploit and resist them in a unified manner simultaneously: 1) Tocombat the negative effects of noisy boundaries, we propose a noise-tolerantmask head by leveraging low-resolution features. 2) To enhance the positiveimpacts, we introduce a boundary-preserving map for learning detailedinformation within boundary-relevant regions. We evaluate our approach byextensive experiments. It behaves extraordinarily, outperforming the supervisedbaseline by a large margin, more than 6% on Cityscapes, 7% on COCO and 4.5% onBDD100k. On Cityscapes, our method achieves comparable performance by utilizingonly 30% labeled images.

Noisy Boundaries: Lemon or Lemonade for Semi-supervised Instance Segmentation?