AI RESEARCH
Structured State-Space Regularization for Compact and Generation-Friendly Image Tokenization
arXiv CS.CV
•
ArXi:2604.11089v1 Announce Type: new Image tokenizers are central to modern vision models as they often operate in latent spaces. An ideal latent space must be simultaneously compact and generation-friendly: it should capture image's essential content compactly while remaining easy to model with generative approaches. In this work, we