推理阶段扩散模型蒸馏
Inference-Time Diffusion Model Distillation
AI总结：
针对扩散蒸馏模型与预训练模型存在性能差距的问题，提出推理阶段蒸馏框架Distillation++，将学生采样重构为带SDS损失的近端优化问题，通过教师引导实时改善去噪，无需额外数据或微调即可显著提升性能。
摘要：
扩散蒸馏模型通过将逆采样过程压缩为更少的步数，有效加速了逆采样过程。然而，与预训练扩散模型相比，这类模型仍存在性能差距，多步采样中的分布偏移与误差累积会进一步拉大这一差距。为解决该问题，我们提出Distillation++，这是一种新型推理阶段蒸馏框架，通过在采样过程中引入教师引导式优化来缩小性能差距。受条件采样领域最新进展的启发，我们的方法将学生模型采样重构为一个带分数蒸馏采样损失（SDS）的近端优化问题。为此，我们在逆采样过程中整合了蒸馏优化，这一操作可被视为教师引导——利用预训练扩散模型推动学生采样轨迹向干净流形靠拢。因此，Distillation++无需额外源数据或微调，即可实时改善去噪过程。Distillation++相较现有最先进的蒸馏基线方法实现了显著提升，尤其在采样早期阶段效果突出，是专为扩散蒸馏模型设计的鲁棒引导采样流程。代码详见：https://github.com/geonyeong-park/inference_distillation。
英文摘要：
Diffusion distillation models effectively accelerate reverse sampling by compressing the process into fewer steps. However, these models still exhibit a performance gap compared to their pre-trained diffusion model counterparts, exacerbated by distribution shifts and accumulated errors during multi-step sampling. To address this, we introduce Distillation++, a novel inference-time distillation framework that reduces this gap by incorporating teacher-guided refinement during sampling. Inspired by recent advances in conditional sampling, our approach recasts student model sampling as a proximal optimization problem with a score distillation sampling loss (SDS). To this end, we integrate distillation optimization during reverse sampling, which can be viewed as teacher guidance that drives student sampling trajectory towards the clean manifold using pre-trained diffusion models. Thus, Distillation++ improves the denoising process in real-time without additional source data or fine-tuning. Distillation++ demonstrates substantial improvements over state-of-the-art distillation baselines, particularly in early sampling stages, positioning itself as a robust guided sampling process crafted for diffusion distillation models. Code: https://github.com/geonyeong-park/inference_distillation.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
