1 paper · 1 filter
Junqiu Yu, Pandeng Li, Yikai Wang +11
Visual tokenizers increasingly inject semantic supervision into latent spaces to make downstream diffusion easier. Yet how these semantics should be organized to facilitate denoisi…