Exploring MLLM-Diffusion Information Transfer with MetaCanvas
By: Han Lin , Xichen Pan , Ziqi Huang and more
Potential Business Impact:
Lets AI draw pictures with better control.
Multimodal learning has rapidly advanced visual understanding, largely via multimodal large language models (MLLMs) that use powerful LLMs as cognitive cores. In visual generation, however, these powerful core models are typically reduced to global text encoders for diffusion models, leaving most of their reasoning and planning ability unused. This creates a gap: current multimodal LLMs can parse complex layouts, attributes, and knowledge-intensive scenes, yet struggle to generate images or videos with equally precise and structured control. We propose MetaCanvas, a lightweight framework that lets MLLMs reason and plan directly in spatial and spatiotemporal latent spaces and interface tightly with diffusion generators. We empirically implement MetaCanvas on three different diffusion backbones and evaluate it across six tasks, including text-to-image generation, text/image-to-video generation, image/video editing, and in-context video generation, each requiring precise layouts, robust attribute binding, and reasoning-intensive control. MetaCanvas consistently outperforms global-conditioning baselines, suggesting that treating MLLMs as latent-space planners is a promising direction for narrowing the gap between multimodal understanding and generation.
Similar Papers
Canvas-to-Image: Compositional Image Generation with Multimodal Controls
CV and Pattern Recognition
Creates pictures from many instructions at once.
How Multimodal LLMs Solve Image Tasks: A Lens on Visual Grounding, Task Reasoning, and Answer Decoding
CV and Pattern Recognition
Shows how AI understands pictures and words.
InstructX: Towards Unified Visual Editing with MLLM Guidance
CV and Pattern Recognition
Edits pictures and videos with simple instructions.