PulseAugur
LIVE 10:12:41
research · [4 sources] ·
0
research

New methods CAGE-SGG and ReLIC-SGG improve open-vocabulary scene graph generation

By PulseAugur Editorial · Summary by gemini-2.5-flash-lite from 4 sources

Two new research papers, CAGE-SGG and ReLIC-SGG, propose novel methods for open-vocabulary scene graph generation. CAGE-SGG focuses on verifying predicted relations using counterfactual evidence to ensure they are visually grounded rather than relying on language priors. ReLIC-SGG addresses the issue of incomplete annotations by treating unannotated relations as latent variables and building a semantic relation lattice to infer missing connections. AI

Summary written by gemini-2.5-flash-lite from 4 sources. How we write summaries →

IMPACT Introduces new techniques for more reliable and interpretable visual scene understanding, potentially improving downstream AI applications.

RANK_REASON The cluster contains two new academic papers on arXiv detailing novel methods for scene graph generation.

Read on arXiv cs.CV →

COVERAGE [4]

  1. arXiv cs.CV TIER_1 · Suiyang Guang, Chenyu Liu, Ruohan Zhang, Siyuan Chen ·

    CAGE-SGG: Counterfactual Active Graph Evidence for Open-Vocabulary Scene Graph Generation

    arXiv:2604.22274v1 Announce Type: new Abstract: Open-vocabulary scene graph generation (SGG) aims to describe visual scenes with flexible and fine-grained relation phrases beyond a fixed predicate vocabulary. While recent vision-language models greatly expand the semantic coverag…

  2. arXiv cs.CV TIER_1 · Amir Hosseini, Sara Farahani, Xinyi Li, Suiyang Guang ·

    ReLIC-SGG: Relation Lattice Completion for Open-Vocabulary Scene Graph Generation

    arXiv:2604.22546v1 Announce Type: new Abstract: Open-vocabulary scene graph generation (SGG) aims to describe visual scenes with flexible relation phrases beyond a fixed predicate set. Existing methods usually treat annotated triplets as positives and all unannotated object-pair …

  3. arXiv cs.CV TIER_1 · Suiyang Guang ·

    ReLIC-SGG: Relation Lattice Completion for Open-Vocabulary Scene Graph Generation

    Open-vocabulary scene graph generation (SGG) aims to describe visual scenes with flexible relation phrases beyond a fixed predicate set. Existing methods usually treat annotated triplets as positives and all unannotated object-pair relations as negatives. However, scene graph ann…

  4. arXiv cs.CV TIER_1 · Siyuan Chen ·

    CAGE-SGG: Counterfactual Active Graph Evidence for Open-Vocabulary Scene Graph Generation

    Open-vocabulary scene graph generation (SGG) aims to describe visual scenes with flexible and fine-grained relation phrases beyond a fixed predicate vocabulary. While recent vision-language models greatly expand the semantic coverage of SGG, they also introduce a critical reliabi…

RELATED ENTITIES

RELATED TOPICS