用于逆成像问题的可逆 ResNet:具有可证明正则化性质的竞争性性能
Invertible ResNets for Inverse Imaging Problems: Competitive Performance with Provable Regularization Properties
AI总结:
本文在真实线性模糊与非线性扩散成像任务上评估 iResNets,证明其性能可媲美先进神经网络,并兼具稳定性、可逆性、鲁棒性与可解释性。
摘要:
基于学习的方法在求解逆问题方面表现出卓越性能,尤其是在图像重建任务中。尽管取得了成功,这些方法通常缺乏理论保证,而这在医学成像等敏感应用中至关重要。Arndt 等人近期的工作通过分析一种基于可逆残差网络(iResNets)的数据驱动重建方法,弥补了这一差距。他们揭示,在合理假设下,该方法构成一种收敛的正则化方案。然而,该重建方法的性能仅在学术性玩具问题和小规模 iResNet 架构上得到验证。本文中,我们通过在两个真实成像任务上评估 iResNets 的性能来解决这一差距:线性模糊算子和非线性扩散算子。为此,我们将 iResNets 的性能与最先进的神经网络进行比较,揭示其具有竞争力,但代价是更长的训练时间。此外,我们通过展示在多种场景下增强的鲁棒性以及所学算子的可解释性,以数值方式证明了 iResNet 固有稳定性和可逆性的优势,从而降低了该重建方案的黑箱性质。
英文摘要:
Learning-based methods have demonstrated remarkable performance in solving inverse problems, particularly in image reconstruction tasks. Despite their success, these approaches often lack theoretical guarantees, which are crucial in sensitive applications such as medical imaging. Recent works by Arndt et al addressed this gap by analyzing a data-driven reconstruction method based on invertible residual networks (iResNets). They revealed that, under reasonable assumptions, this approach constitutes a convergent regularization scheme. However, the performance of the reconstruction method was only validated on academic toy problems and small-scale iResNet architectures. In this work, we address this gap by evaluating the performance of iResNets on two real-world imaging tasks: a linear blurring operator and a nonlinear diffusion operator. To do so, we compare the performance of iResNets against state-of-the-art neural networks, revealing their competitiveness at the expense of longer training times. Moreover, we numerically demonstrate the advantages of the iResNet's inherent stability and invertibility by showcasing increased robustness across various scenarios as well as interpretability of the learned operator, thereby reducing the black-box nature of the reconstruction scheme.
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。