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/

Zendy Plus

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

Zendy Tools

Zendy Open

Instance segmentation in remote sensing images is a long-standing challenge.Since horizontal bounding boxes introduce many interference objects, orientedbounding boxes (OBBs) are usually used for instance identification. However,based on ``segmentation within bounding box'' paradigm, current instancesegmentation methods using OBBs are overly dependent on bounding box detectionperformance. To tackle this problem, this paper proposes OBSeg, an accurate andfast instance segmentation framework using OBBs. OBSeg is based on boxprompt-based segmentation foundation models (BSMs), e.g., Segment AnythingModel. Specifically, OBSeg first detects OBBs to distinguish instances andprovide coarse localization information. Then, it predicts OBB prompt-relatedmasks for fine segmentation. Since OBBs only serve as prompts, OBSeg alleviatesthe over-dependence on bounding box detection performance of current instancesegmentation methods using OBBs. Thanks to OBB prompts, OBSeg outperforms othercurrent BSM-based methods using HBBs. In addition, to enable BSMs to handle OBBprompts, we propose a novel OBB prompt encoder. To make OBSeg more lightweightand further improve the performance of lightweight distilled BSMs, a Gaussiansmoothing-based knowledge distillation method is introduced. Experimentsdemonstrate that OBSeg outperforms current instance segmentation methods onmultiple datasets in terms of instance segmentation accuracy and hascompetitive inference speed. The code is available athttps://github.com/zhen6618/OBBInstanceSegmentation.

OBSeg: Accurate and Fast Instance Segmentation Framework Using Segmentation Foundation Models with Oriented Bounding Box Prompts