通过二维并行解码，让自回归图像生成快4.8到6.4倍。
Parallel Jacobi Decoding for Fast Autoregressive Image Generation
- 在二维空间并行扩展候选像素，打破一维序列依赖
- 实测加速4.8至6.4倍，图像质量接近原模型
- 无需训练，适配多种自回归图像模型
自回归（AR）模型在生成高保真图像方面表现卓越，但其固有的逐词预测机制导致推理速度缓慢。近期研究引入类似雅可比的解码方式以加速生成过程。尽管初始扩展候选序列能提升效率，但随着一维序列中误差传播加剧，加速效果迅速饱和。鉴于图像具有强局部空间相关性，我们提出无需训练的并行雅可比解码（PJD），在二维空间域扩展候选像素，实现高效的空间并行优化。PJD通过调整注意力掩码缓解误差累积，提升收敛稳定性。在多个数据集上的大量实验表明，PJD在多种自回归图像生成模型上实现了4.8至6.4倍的加速，同时保持优异的生成质量。
原文摘要 · Abstract (English)
Autoregressive (AR) models have demonstrated remarkable performance in generating high-fidelity images. However, their inherently sequential next-token prediction leads to significantly slower inference. Recent studies have introduced Jacobi-style decoding to accelerate autoregressive image generation. Extending the draft sequence initially improves efficiency, yet the acceleration quickly saturates as error propagation in the one-dimensional sequence hinders convergence. Observing that images exhibit strong local spatial correlations, we propose Parallel Jacobi Decoding (PJD), a training-free decoding approach that expands draft tokens in the two-dimensional spatial domain to enable efficient spatially parallel refinement. PJD adjusts the attention mask to mitigate error accumulation and improve convergence stability. Extensive experiments on diverse datasets show that PJD achieves 4.8x-6.4x acceleration across multiple autoregressive image generation models while maintaining competitive generation quality.
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。