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

Recently, semantic segmentation models trained with image-level textsupervision have shown promising results in challenging open-world scenarios.However, these models still face difficulties in learning fine-grained semanticalignment at the pixel level and predicting accurate object masks. To addressthis issue, we propose MixReorg, a novel and straightforward pre-trainingparadigm for semantic segmentation that enhances a model's ability toreorganize patches mixed across images, exploring both local visual relevanceand global semantic coherence. Our approach involves generating fine-grainedpatch-text pairs data by mixing image patches while preserving thecorrespondence between patches and text. The model is then trained to minimizethe segmentation loss of the mixed images and the two contrastive losses of theoriginal and restored features. With MixReorg as a mask learner, conventionaltext-supervised semantic segmentation models can achieve highly generalizablepixel-semantic alignment ability, which is crucial for open-world segmentation.After training with large-scale image-text data, MixReorg models can be applieddirectly to segment visual objects of arbitrary categories, without the needfor further fine-tuning. Our proposed framework demonstrates strong performanceon popular zero-shot semantic segmentation benchmarks, outperforming GroupViTby significant margins of 5.0%, 6.2%, 2.5%, and 3.4% mIoU on PASCAL VOC2012,PASCAL Context, MS COCO, and ADE20K, respectively.

MixReorg: Cross-Modal Mixed Patch Reorganization is a Good Mask Learner for Open-World Semantic Segmentation