用可变长度令牌实现任意分辨率图像生成，效率远超扩散模型。
VibeToken: Scaling 1D Image Tokenizers and Autoregressive Models for Dynamic Resolution Generations
- 1D Transformer tokenizer 动态生成32-256个令牌，适配任意分辨率。
- 1024x1024图像仅需64令牌，生成质量gFID达3.94，优于扩散模型。
- 推理计算量恒定179G FLOPs，是传统自回归模型的63.4倍效率。
我们提出一种高效、分辨率无关的自回归（AR）图像生成方法，能泛化至任意分辨率与长宽比，显著缩小与扩散模型在大规模下的差距。核心为VibeToken，一种基于1D Transformer的新型分辨率无关图像分词器，将图像编码为动态可控的32-256个令牌序列，实现了当前最优的效率与性能权衡。在此基础上，我们构建了支持任意分辨率的类条件自回归生成器VibeToken-Gen，其推理资源消耗极低。例如，生成1024x1024图像仅需64令牌，达到3.94 gFID；而同类扩散模型需1024令牌，仅得5.87 gFID。相较固定分辨率的自回归模型如LlamaGen——其推理浮点运算量随分辨率平方增长（1024x1024时达11T FLOPs）——VibeToken-Gen保持恒定179G FLOPs（效率提升63.4倍），不随分辨率变化。我们希望VibeToken推动自回归视觉生成模型在生产环境中的广泛应用。
原文摘要 · Abstract (English)
We introduce an efficient, resolution-agnostic autoregressive (AR) image synthesis approach that generalizes to arbitrary resolutions and aspect ratios, narrowing the gap to diffusion models at scale. At its core is VibeToken, a novel resolution-agnostic 1D Transformer-based image tokenizer that encodes images into a dynamic, user-controllable sequence of 32-256 tokens, achieving a state-of-the-art efficiency and performance trade-off. Building on VibeToken, we present VibeToken-Gen, a class-conditioned AR generator with out-of-the-box support for arbitrary resolutions while requiring significantly fewer compute resources. Notably, VibeToken-Gen synthesizes 1024x1024 images using only 64 tokens and achieves 3.94 gFID; by comparison, a diffusion-based state-of-the-art alternative requires 1,024 tokens and attains 5.87 gFID. In contrast to fixed-resolution AR models such as LlamaGen -- whose inference FLOPs grow quadratically with resolution (11T FLOPs at 1024x1024) -- VibeToken-Gen maintains a constant 179G FLOPs (63.4x efficient) independent of resolution. We hope VibeToken can help unlock the wide adoption of AR visual generative models in production use cases.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。