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

Modern object detectors have achieved impressive progress under the close-setsetup. However, open-set object detection (OSOD) remains challenging sinceobjects of unknown categories are often misclassified to existing knownclasses. In this work, we propose to identify unknown objects by separatinghigh/low-density regions in the latent space, based on the consensus thatunknown objects are usually distributed in low-density latent regions. Astraditional threshold-based methods only maintain limited low-density regions,which cannot cover all unknown objects, we present a novel Open-set Detector(OpenDet) with expanded low-density regions. To this aim, we equip OpenDet withtwo learners, Contrastive Feature Learner (CFL) and Unknown Probability Learner(UPL). CFL performs instance-level contrastive learning to encourage compactfeatures of known classes, leaving more low-density regions for unknownclasses; UPL optimizes unknown probability based on the uncertainty ofpredictions, which further divides more low-density regions around the clusterof known classes. Thus, unknown objects in low-density regions can be easilyidentified with the learned unknown probability. Extensive experimentsdemonstrate that our method can significantly improve the OSOD performance,e.g., OpenDet reduces the Absolute Open-Set Errors by 25%-35% on six OSODbenchmarks. Code is available at: https://github.com/csuhan/opendet2.

Expanding Low-Density Latent Regions for Open-Set Object Detection