Metraipolis › archive › dc_8cd54985d630

INTCORT: Training-Free Spatial Reasoning Enhancement for Vision-Language Models via Input Transformations and Confidence Routing — arXiv:2609.24813

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.24813
License (as archived)
CC0-1.0
Archived
2026-10-09T22:06:23.680Z
SHA-256 of the archived text
55d72ce4c7d2cfd11990de6db740423229843d08c05f35a1701fb31139608653
Stable URL of this page
https://metraipolis.com/documento/dc_8cd54985d630

How to cite

INTCORT: Training-Free Spatial Reasoning Enhancement for Vision-Language Models via Input Transformations and Confidence Routing — arXiv:2609.24813 — arXiv (en), https://arxiv.org/abs/2609.24813 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_8cd54985d630 (Metraipolis archive dc_8cd54985d630, sha256 55d72ce4c7d2cfd11990de6db740423229843d08c05f35a1701fb31139608653)
BibTeX
@misc{dc8cd54985d630,
  title = {INTCORT: Training-Free Spatial Reasoning Enhancement for Vision-Language Models via Input Transformations and Confidence Routing — arXiv:2609.24813},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.24813},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_8cd54985d630; sha256 55d72ce4c7d2cfd11990de6db740423229843d08c05f35a1701fb31139608653},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.24813. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Vision-Language Models (VLMs) have demonstrated remarkable capabilities in multimodal tasks, yet they still exhibit poor ability in spatial reasoning. Existing training-dependent and training-free enhancement methods suffer from high computational costs with catastrophic forgetting and internal mechanism interference that compromises general capabilities, respectively. In this work, we first verify two key hypotheses: appropriate geometric image transformation and query-reversal transformation can recover incorrect spatial predictions, and correct predictions exhibit higher relation-token confidence than incorrect ones. Based on these findings, we propose INTCORT, a training-free spatial reasoning enhancement framework that constructs multiple inference views through input transformations and aggregates their predictions via relation-token confidence routing, without modifying the VLM's internal mechanisms. Experimental results on several commonly-used benchmarks demonstrate that INTCORT substantially improves spatial reasoning accuracy across diverse VLMs, achieving an average improvement of 10.01% over all models and benchmarks. Compared with prior works, INTCORT achieves superior performance with improvements of up to 25.01%.

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-22 in cs.CV).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2609.24813 — 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 2026-10-09T21:45:21.094Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-09-22?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).