UniCanvas: diffusion-based unified model for text-in-image generation

Arxiv CS.CV (computer vision)

UniCanvas is a diffusion-based unified model for joint text and image generation. Unlike autoregressive VLMs, it handles both multimodal understanding and generation within a single architecture.

diffusionmultimodal

Arxiv CS.CV (computer vision)·Zeyuan YangHao-Wei ChenXueyang YuYuncong YangHaoyu ZhenZiqiao MaMaohao ShenChuang Gan

7 days ago