AI RESEARCH
Cross-Domain Lossy Compression via Constrained Minimum Entropy Coupling
arXiv CS.LG
•
ArXi:2605.09833v1 Announce Type: cross This paper studies cross-domain lossy compression through the lens of minimum entropy coupling (MEC) with rate and classification constraints. In this setting, an encoder observes samples from a degraded source domain, while the decoder is required to generate outputs following a prescribed target distribution and to preserve information relevant to a downstream classification task.