AI Paper Atlas

Research map / Vision-language and multimodal / Cross-modal learning

SphereFlow: Missing Modality Imputation via Geometric Transport on Hypersphere

Huihan Wang, Juncheng Wang, Yuxiang Feng, Wenlong Hou, Guoqi Yu, Chao Xu, Yang Liu, Baigui Sun, Shujun Wang Emma

NeurIPS 2026 · Vision-language and multimodal › Cross-modal learning · approach: Multi-modality image fusion · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes SphereFlow, which imputes missing modalities by transporting latents along geometry on a shared hyperspherical space obtained from a frozen self-supervised encoder.

Problem

Missing modality imputation: synthesizing unobserved modalities from observed ones. Existing methods treat each modality as on its own latent manifold, forcing the generator to learn a long unstructured jump between modalities.

Approach

Projects all modalities into a shared hyperspherical latent space via a frozen self-supervised encoder so cross-modal imputation becomes geometric transport on the hypersphere rather than an unstructured jump between per-modality manifolds.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
replace Gaussian noise source with observed-modality latent and add manifold-expanded decoder; 1-to-1 BraTSSphereFlow · BraTS 1-to-1rFID: 29.99 -> 19.93Improved
SphereFlow vs best baseline on downstream segmentationSphereFlow · BraTS segmentation avgDice: 0.5930 -> 0.7473Improved
use DINOv2 encoder instead of MedMAE (ablation)SphereFlow · Stage-1 T1c reconstructioniFID: 144.45 -> 40.64Improved
use DINOv2 encoder instead of MedMAE (ablation)SphereFlow · Stage-1 T1c reconstructionPSNR: 31.23 -> 30.57Hurt
add Step 2 manifold expansion to decoder (ablation)SphereFlow RAE · on-manifoldPSNR: 33.14 -> 33.83Improved
add Step 2 manifold expansion to decoder (ablation)SphereFlow RAE · on-manifoldrFID: 28.29 -> 30.92Hurt
add Step 3 shared-encoder alignment (ablation)SphereFlow RAE · on-manifoldPSNR: 33.83 -> 44.06Improved
SphereFlow vs Zhang on CT to MRISphereFlow · CT to MRIPSNR: 25.51 -> 24.70Hurt
SphereFlow vs Zhang on CT to MRISphereFlow · CT to MRIrFID: 30.91 -> 13.65Improved
average latents of multiple source modalities (N-to-N)SphereFlow · BraTS N-to-NPSNR: 35.13 -> 35.88Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Experimental setup

Builds on: flow matching

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy