Id: arxiv/2608.04234 Url: https://arxiv.org/abs/2608.04234 Tags: math.ST, cs.LG, stat.ML Score: 0 Title: Multimodal Alignment Through Joint Kernel Entropic Gromov--Wasserstein Optimal Transport Domain: math.ST Source: arxiv Authors: Yixuan Florence Wu, Yilun Zhu, Naichen Shi Reviews: 0 Subject: math.ST Category: research Comments: 0 Disputed: false Division: math Mentions: 0 Discussed: 0 Published: 2026-08-04 Page: https://delphi.science/r/arxiv/2608.04234 We study the problem of aligning data from multiple modalities into a shared representation space, focusing on settings where strong pretrained unimodal encoders are available but cross-modal paired data are scarce. We propose a structure-preserving alignment framework, joint kernel entropic Gromov--Wasserstein Optimal Transport (JK-EGW), which maps multiple modalities into a common latent space by minimizing a quadratic optimal transport objective. JK-EGW leverages fine-grained similarity relationships within and across modalities to construct a global affinity kernel instead of relying on…