Metraipolis › archive › dc_e392e003a07a

ConstraintMatch: Compositional Image-Text Matching through Constrained Regional Evidence Selection — arXiv:2611.00784

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.00784
License (as archived)
CC0-1.0
Archived
2026-11-03T18:24:33.254Z
SHA-256 of the archived text
4cb7e1668f9bfbf91f3fe06b285c53cf2b443467efb64b15b72cbd7429e65fbd
Stable URL of this page
https://metraipolis.com/documento/dc_e392e003a07a

How to cite

ConstraintMatch: Compositional Image-Text Matching through Constrained Regional Evidence Selection — arXiv:2611.00784 — arXiv (en), https://arxiv.org/abs/2611.00784 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_e392e003a07a (Metraipolis archive dc_e392e003a07a, sha256 4cb7e1668f9bfbf91f3fe06b285c53cf2b443467efb64b15b72cbd7429e65fbd)
BibTeX
@misc{dce392e003a07a,
  title = {ConstraintMatch: Compositional Image-Text Matching through Constrained Regional Evidence Selection — arXiv:2611.00784},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.00784},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_e392e003a07a; sha256 4cb7e1668f9bfbf91f3fe06b285c53cf2b443467efb64b15b72cbd7429e65fbd},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.00784. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Local image--text matching can find strong evidence for individual concepts yet favor captions with incorrect attribute--entity bindings. We introduce ConstraintMatch, a training-free inference method that turns entity-level regional consistency into an explicit matching cost. On the same frozen condition--region similarity matrix, it compares independent crop selection for each category and attribute with a shared crop choice within each entity. Their gap quantifies the score advantage obtained by relaxing the shared-region constraint. Spatial references from the same image calibrate this gap by keeping category evidence fixed while jointly shifting an entity's attribute maps. Per-entity clipping prevents cancellation across residuals, and the scaled mean cost is subtracted from global similarity. ConstraintMatch uses pooled embeddings, and spatial calibration requires no additional encoder calls. With parameters fixed per backbone across tasks, ConstraintMatch improves over global matching by 7.70 percentage points on full ARO Attribute and 5.63 points on held-out SugarCrepe++ attribute swaps, averaged across Qwen~2B and CORE~2B. Against SGI adapted to the same 13-crop budget, the respective gains are 4.85 and 3.51 points. Controlled analyses distinguish the effects of shared-region matching and candidate score calibration. The results establish entity-level evidence selection as a practical scoring principle for compositional multimodal retrieval.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2611.00784 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-11-03T18:10:08.020Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).