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

Humans inherently recognize objects via selective visual perception,transform specific regions from the visual field into structured symbolicknowledge, and reason their relationships among regions based on the allocationof limited attention resources in line with humans' goals. While it isintuitive for humans, contemporary perception systems falter in extractingstructural information due to the intricate cognitive abilities and commonsenseknowledge required. To fill this gap, we present a new task called Local SceneGraph Generation. Distinct from the conventional scene graph generation task,which encompasses generating all objects and relationships in an image, ourproposed task aims to abstract pertinent structural information with partialobjects and their relationships for boosting downstream tasks that demandadvanced comprehension and reasoning capabilities. Correspondingly, weintroduce zEro-shot Local scEne GrAph geNeraTion (ELEGANT), a frameworkharnessing foundation models renowned for their powerful perception andcommonsense reasoning, where collaboration and information communication amongfoundation models yield superior outcomes and realize zero-shot local scenegraph generation without requiring labeled supervision. Furthermore, we proposea novel open-ended evaluation metric, Entity-level CLIPScorE (ECLIPSE),surpassing previous closed-set evaluation metrics by transcending their limitedlabel space, offering a broader assessment. Experiment results show that ourapproach markedly outperforms baselines in the open-ended evaluation setting,and it also achieves a significant performance boost of up to 24.58% over priormethods in the close-set setting, demonstrating the effectiveness and powerfulreasoning ability of our proposed framework.

Less is More: Toward Zero-Shot Local Scene Graph Generation via Foundation Models