AI RESEARCH
Domain Transfer Becomes Identifiable via a Single Alignment
arXiv CS.LG
•
ArXi:2605.17918v1 Announce Type: new Domain transfer (DT) maps source to target distributions and s tasks such as unsupervised image-to-image translation, single-cell analysis, and cross-platform medical imaging. However, DT is fundamentally ill-posed: push-forward mappings are generally non-identifiable, as measure-preserving automorphisms (MPAs) preserve marginals while altering cross-domain correspondences, leading to content-misaligned translation.