发表机构
Sichuan University; School of Artificial Intelligence, Sichuan University（四川大学; 四川大学人工智能学院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究针对扩散Transformer（DiTs）高加速推理时的质量下降问题，提出带切比雪夫增强的重心有理预测方法BRACE，通过特征驱动的有理预测实现最优质量-效率权衡，且计算开销极低。
AI 中文摘要
扩散Transformer（DiTs）在高保真图像和视频生成中展现出卓越性能。为缓解其巨大计算开销，研究人员提出了时间特征缓存以绕过冗余计算。然而，现有基于导数多项式的缓存后预测方法在高加速场景下，因长步预测不稳定常导致严重质量下降。为解决这一瓶颈，我们提出带切比雪夫增强的重心有理预测方法BRACE。鉴于DiT特征轨迹全局平滑但常出现尖锐不规则性和局部非平滑性，BRACE将范式从导数驱动的多项式外推转向特征驱动的有理预测。具体而言，它维护局部滑动窗口以缓存稀疏历史特征，并利用适配的切比雪夫权重构建重心有理函数，直接聚合这些原始特征以确保数值稳定性。大量实验表明，BRACE在各类DiT架构上实现了最优的质量-效率权衡，且计算开销可忽略不计。
英文摘要
Diffusion Transformers (DiTs) have demonstrated exceptional performance in high-fidelity image and video generation. To alleviate their massive computational overhead, temporal feature caching has been proposed to bypass redundant computations. However, existing cache-then-forecast methods driven by derivative-based polynomials often cause severe quality degradation under high acceleration due to unstable long-step predictions. To address this bottleneck, we propose Barycentric Rational Forecasting with Chebyshev Enhancement (BRACE). Motivated by the observation that DiT feature trajectories are globally smooth yet frequently exhibit sharp irregularities and local non-smoothness, BRACE shifts the paradigm from derivative-driven polynomial extrapolation to feature-driven rational forecasting. Specifically, it maintains a local sliding window to cache sparse historical features and leverages adapted Chebyshev weights to formulate a barycentric rational function, directly aggregating these raw features to ensure numerical stability. Extensive experiments demonstrate that BRACE achieves state-of-the-art quality-efficiency trade-offs across various DiT architectures with negligible computational overhead.
CommentsAccepted at ACM MM 2026. Project page: https://youngkinlon.github.io/BRACE-Taming-Sharp-Irregularities-via-Barycentric-Rational-Forecasting-for-Fast-DiT-Inference/
