Score: 2

Decoupling Complexity from Scale in Latent Diffusion Model

Published: November 20, 2025 | arXiv ID: 2511.16117v1

By: Tianxiong Zhong , Xingye Tian , Xuebo Wang and more

BigTech Affiliations: Kuaishou

Potential Business Impact:

Makes pictures and videos with any detail.

Business Areas:
Image Recognition Data and Analytics, Software

Existing latent diffusion models typically couple scale with content complexity, using more latent tokens to represent higher-resolution images or higher-frame rate videos. However, the latent capacity required to represent visual data primarily depends on content complexity, with scale serving only as an upper bound. Motivated by this observation, we propose DCS-LDM, a novel paradigm for visual generation that decouples information complexity from scale. DCS-LDM constructs a hierarchical, scale-independent latent space that models sample complexity through multi-level tokens and supports decoding to arbitrary resolutions and frame rates within a fixed latent representation. This latent space enables DCS-LDM to achieve a flexible computation-quality tradeoff. Furthermore, by decomposing structural and detailed information across levels, DCS-LDM supports a progressive coarse-to-fine generation paradigm. Experimental results show that DCS-LDM delivers performance comparable to state-of-the-art methods while offering flexible generation across diverse scales and visual qualities.

Country of Origin
🇨🇳 China

Page Count
15 pages

Category
Computer Science:
CV and Pattern Recognition