按复杂度从简到繁训练扩散模型，显著提升效率与质量。
Data Warmup: Complexity-Aware Curricula for Efficient Diffusion Training
- 用语义感知复杂度评分筛选图像，从简单到复杂逐步训练。
- 在ImageNet上使IS提升6.11、FID降低3.41，提前数万步达基线效果。
- 无需修改模型或损失函数，仅需10分钟预处理，适合加速扩散训练。
扩散模型训练中的关键低效问题在于：随机初始化的网络缺乏视觉先验，却要应对全范围复杂度的梯度，多数难以处理。我们提出Data Warmup，一种不修改模型或损失的课程学习策略。通过离线计算每张图像的复杂度得分——结合前景主导性（显著物体占画面比例）和前景典型性（显著内容与已学视觉原型的匹配度），再用温度控制采样器优先选择低复杂度图像，随训练进程逐渐转向均匀采样。在使用SiT主干网络（S/2至XL/2）的ImageNet 256x256数据集上，该方法使IS提升最高达6.11，FID降低最多3.41，提前数万次迭代达到基线性能。反向课程（先训难图）表现低于均匀采样基线，证实了从简到繁顺序的有效性。该方法可与REPA等加速技术兼容，仅需约10分钟一次性预处理，无每次迭代开销。
原文摘要 · Abstract (English)
A key inefficiency in diffusion training occurs when a randomly initialized network, lacking visual priors, encounters gradients from the full complexity spectrum--most of which it lacks the capacity to resolve. We propose Data Warmup, a curriculum strategy that schedules training images from simple to complex without modifying the model or loss. Each image is scored offline by a semantic-aware complexity metric combining foreground dominance (how much of the image salient objects occupy) and foreground typicality (how closely the salient content matches learned visual prototypes). A temperature-controlled sampler then prioritizes low-complexity images early and anneals toward uniform sampling. On ImageNet 256x256 with SiT backbones (S/2 to XL/2), Data Warmup improves IS by up to 6.11 and FID by up to 3.41, reaching baseline quality tens of thousands of iterations earlier. Reversing the curriculum (exposing hard images first) degrades performance below the uniform baseline, confirming that the simple-to-complex ordering itself drives the gains. The method combines with orthogonal accelerators such as REPA and requires only ~10 minutes of one-time preprocessing with zero per-iteration overhead.
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。