Metraipolis › archive › dc_74cc3ddd6db5

SphereFlow: Missing Modality Imputation via Geometric Transport on Hypersphere — arXiv:2611.04414

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.04414
License (as archived)
CC0-1.0
Archived
2026-11-04T12:56:23.631Z
SHA-256 of the archived text
5c78a339029f028214b0caf2e3df7e0147924f488cfc1da6e4aeb199e0f6f56f
Stable URL of this page
https://metraipolis.com/documento/dc_74cc3ddd6db5

How to cite

SphereFlow: Missing Modality Imputation via Geometric Transport on Hypersphere — arXiv:2611.04414 — arXiv (en), https://arxiv.org/abs/2611.04414 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_74cc3ddd6db5 (Metraipolis archive dc_74cc3ddd6db5, sha256 5c78a339029f028214b0caf2e3df7e0147924f488cfc1da6e4aeb199e0f6f56f)
BibTeX
@misc{dc74cc3ddd6db5,
  title = {SphereFlow: Missing Modality Imputation via Geometric Transport on Hypersphere — arXiv:2611.04414},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.04414},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_74cc3ddd6db5; sha256 5c78a339029f028214b0caf2e3df7e0147924f488cfc1da6e4aeb199e0f6f56f},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.04414. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Missing modality imputation aims to synthesize unobserved modalities from observed ones. Existing methods treat each modality as living on its own latent manifold, forcing the generator to learn a long, unstructured jump between them. We present SphereFlow, which takes a fundamentally different view: by projecting all modalities into a shared hyperspherical latent space via a frozen self-supervised encoder, cross-modal imputation reduces to short-range geometric transport between nearby points on the sphere. Realizing this idea with flow matching, however, exposes two geometric conflicts: the standard Gaussian noise source is catastrophically mismatched with the unit-norm data manifold, and the linear interpolation path departs the sphere surface into a region where the decoder has no training signal. We resolve both with simple, geometry-aware modifications: replacing the noise source with the observed-modality latent so that the velocity field learns only the cross-modal displacement, and training the decoder with a time-weighted loss that extends its domain to the near-surface region traversed by the flow. We prove that this data-as-source formulation reduces the worst-case off-manifold deviation from a quantity that grows with the latent dimension to a small, dimension-independent bound governed by inter-modality similarity. Experiments on BraTS multi-modal MRI imputation and CT-MRI translation demonstrate state-of-the-art generation fidelity across all settings, while downstream brain tumor segmentation with our imputed data closes roughly eighty percent of the gap to the full-modality oracle. Code is available at this https URL .

Tema degli esploratori: statistica e scienza dei dati (elenco arXiv cs.CV, nuovi lavori del giorno).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2611.04414 — 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 2026-11-04T12:33:47.141Z dalla pagina https://arxiv.org/list/cs.CV/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).