追踪神谕:改进扩散时间步调度用于3D CT重建
Tracing the Oracle: Improving Diffusion Timestep Scheduling for 3D CT Reconstruction
AI总结:
针对3D CT重建中扩散模型推理计算开销大且均匀时间步调度引入大截断误差的问题,提出即插即用的TrO框架,通过动态规划优化时间步调度,在有限采样步数下显著提升重建保真度和计算效率。
摘要:
预训练扩散模型在解决高度病态的3D计算机断层扫描(CT)逆问题中展现出令人印象深刻的潜力,但推理过程存在显著的计算开销。此外,现有的均匀时间步调度未能捕捉反向条件扩散随机微分方程的非均匀演化,从而引入了大量截断误差。为克服这一限制,我们提出Tracing the Oracle(TrO),一种用于改进时间步调度的即插即用框架。具体而言,我们将少量样本上的密集采样数值积分轨迹视为参考神谕。通过动态规划全局最小化少步近似与神谕之间的累积误差,提取优化后的调度。该机制将有限的采样步精确分配到对截断误差高度敏感的关键演化阶段。我们在AAPM数据集上的多个3D CT重建任务中进行的广泛实验表明,与最先进的3D CT重建方法DDS结合时,我们的优化时间步在不超过10个采样步的严格预算下,相比现有启发式调度显著提高了重建保真度和计算效率。
英文摘要:
Pretrained diffusion models demonstrate impressive potential in solving highly ill-posed 3D computed tomography (CT) inverse problems, while the inference process suffers from significant computational overhead. Furthermore, existing uniform timestep schedules fail to capture the non-uniform evolution of the reverse conditional diffusion stochastic differential equation, thereby introducing substantial truncation errors. To overcome this limitation, we propose Tracing the Oracle (TrO), a plug-and-play framework for improved timestep scheduling. Specifically, we treat densely sampled numerical integration trajectories on a few samples as the reference oracle. The optimized schedule is extracted by leveraging dynamic programming to globally minimize the cumulative error between the few-step approximation and the oracle. This mechanism precisely allocates the limited sampling steps to critical evolution stages that are highly susceptible to truncation errors. Our extensive experiments on the AAPM dataset across multiple 3D CT reconstruction tasks demonstrate that, when combined with the state-of-the-art 3D CT reconstruction method DDS, our optimized timesteps significantly improve reconstruction fidelity and computational efficiency compared to existing heuristic schedules, especially under a strict budget of no more than 10 sampling steps.
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。