Metraipolis › archive › dc_a63fade8cc58

Do More Modalities Always Help? A Geometric Perspective on Missing-Modality Robustness — arXiv:2610.04792

Source
arXiv
Original
https://arxiv.org/abs/2610.04792
License (as archived)
CC0-1.0
Archived
2026-10-06T05:24:47.923Z
SHA-256 of the archived text
9431fb6dc887c5c98517e14be65558a2cc573e9149da8765ef80217470ffc638
Stable URL of this page
https://metraipolis.com/documento/dc_a63fade8cc58

How to cite

Do More Modalities Always Help? A Geometric Perspective on Missing-Modality Robustness — arXiv:2610.04792 — arXiv, https://arxiv.org/abs/2610.04792 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_a63fade8cc58 (Metraipolis archive dc_a63fade8cc58, sha256 9431fb6dc887c5c98517e14be65558a2cc573e9149da8765ef80217470ffc638)
BibTeX
@misc{dca63fade8cc58,
  title = {Do More Modalities Always Help? A Geometric Perspective on Missing-Modality Robustness — arXiv:2610.04792},
  howpublished = {arXiv},
  url = {https://arxiv.org/abs/2610.04792},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_a63fade8cc58; sha256 9431fb6dc887c5c98517e14be65558a2cc573e9149da8765ef80217470ffc638},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv, https://arxiv.org/abs/2610.04792. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Missing modality remains a longstanding challenge in multimodal learning. Existing methods typically address this issue through modality recovery or adaptive strategies. However, they overlook models' internal cross-modal dependencies formed during multimodal training, which later impair robustness. We systematically characterize a counterintuitive deployment-time failure mode: models trained on full modalities can underperform unimodal models when one modality is missing at inference time. This pattern appears across diverse architectures, such as fusion models, CLIP-style two-tower models, and vision-language models. We show that such degradation is closely associated with learned cross-modal dependencies in the principal parameter subspaces. Multimodal training induces structured rotations of these subspaces, particularly in cross-modal interaction layers. These rotations are associated with reduced task-aligned margins and larger task-aware representation harm under missing-modality inputs. We propose Geodesic Unlearning (GU), a lightweight parameter-editing method that leverages Grassmannian subspace geometry for structured subspace correction to improve missing-modality robustness. It rotates the principal input subspace toward a unimodal reference along a geodesic path. We prove that this correction minimizes the distance to the reference within a fixed subspace-distance budget. Experiments across architectures and datasets show that GU improves performance under missing-modality inference while preserving full-modality accuracy, outperforming strong missing-modality robustness baselines. These findings support a geometric view of deployment-time missing-modality degradation and suggest localized subspace editing as a practical route for robustness correction.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Artificial Intelligence (cs.AI) ; Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2610.04792 — 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 2026-10-06T05:24:05.517Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).