{"paper":{"title":"ReLIC-SGG: Relation Lattice Completion for Open-Vocabulary Scene Graph Generation","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"ReLIC-SGG infers missing relations in open-vocabulary scene graphs using a semantic predicate lattice.","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Amir Hosseini, Sara Farahani, Suiyang Guang, Xinyi Li","submitted_at":"2026-04-24T13:36:41Z","abstract_excerpt":"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 annotations are inherently incomplete: many valid relations are missing, and the same interaction can be described at different granularities, e.g., \\textit{on}, \\textit{standing on}, \\textit{resting on}, and \\textit{supported by}. This issue becomes more severe in open-vocabulary SGG due to the much larger relation space"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"ReLIC-SGG improves rare and unseen predicate recognition and better recovers missing relations on conventional, open-vocabulary, and panoptic SGG benchmarks.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the semantic relation lattice plus visual-language compatibility and graph context can reliably distinguish true missing positives from true negatives without introducing more errors than it removes.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"ReLIC-SGG completes missing open-vocabulary scene-graph relations via a semantic lattice and positive-unlabeled learning rather than treating unannotated pairs as negatives.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"ReLIC-SGG infers missing relations in open-vocabulary scene graphs using a semantic predicate lattice.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"d75f684283ca52efe48d4576a80d60fc2fb6aa8579fb37fec04bc9eaeaddcb48"},"source":{"id":"2604.22546","kind":"arxiv","version":6},"verdict":{"id":"421e00e2-53fd-4549-94f8-03488dff2c4d","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-14T21:27:15.145639Z","strongest_claim":"ReLIC-SGG improves rare and unseen predicate recognition and better recovers missing relations on conventional, open-vocabulary, and panoptic SGG benchmarks.","one_line_summary":"ReLIC-SGG completes missing open-vocabulary scene-graph relations via a semantic lattice and positive-unlabeled learning rather than treating unannotated pairs as negatives.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the semantic relation lattice plus visual-language compatibility and graph context can reliably distinguish true missing positives from true negatives without introducing more errors than it removes.","pith_extraction_headline":"ReLIC-SGG infers missing relations in open-vocabulary scene graphs using a semantic predicate lattice."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.22546/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"ai_meta_artifact","ran_at":"2026-05-21T10:37:44.323342Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T23:54:26.675181Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"3876f39c764e2cf57e00f22345c1c70d9aaf0b394bba975227950fb5d46d2700"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":1,"snapshot_sha256":"7b6bf3bfa7376dc83c1664cac1b7614ab3d4be12013936fd7ba6ce1fe0418058"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}