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 Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Open-world instance segmentation has recently gained significantpopularitydue to its importance in many real-world applications, such asautonomous driving, robot perception, and remote sensing. However, previousmethods have either produced unsatisfactory results or relied on complexsystems and paradigms. We wonder if there is a simple way to obtainstate-of-the-art results. Fortunately, we have identified two observations thathelp us achieve the best of both worlds: 1) query-based methods demonstratesuperiority over dense proposal-based methods in open-world instancesegmentation, and 2) learning localization cues is sufficient for open worldinstance segmentation. Based on these observations, we propose a simplequery-based method named OpenInst for open world instance segmentation.OpenInst leverages advanced query-based methods like QueryInst and focuses onlearning localization cues. Notably, OpenInst is an extremely simple andstraightforward framework without any auxiliary modules or post-processing, yetachieves state-of-the-art results on multiple benchmarks. Specifically, in theCOCO$\to$UVO scenario, OpenInst achieves a mask AR of 53.3, outperforming theprevious best methods by 2.0 AR with a simpler structure. We hope that OpenInstcan serve as a solid baselines for future research in this area.

OpenInst: A Simple Query-Based Method for Open-World Instance Segmentation