arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~

QuScope:用于透射电子显微镜量子电路模拟的开源Python框架

QuScope: An Open-Source Python Framework for Quantum-Circuit Simulation of Transmission Electron Microscopy

Sean D. Lam, Roberto dos Reis

arXiv 2608.02782首次发表:更新:

AI 中文总结

提出开源Python框架QuScope,将透射电镜成像流程建模为量子电路,实现了经验证的成像流程,其结果与经典方法一致,还提供量子资源估计及相关分析。

AI 中文摘要

透射电子显微镜(TEM)中的图像形成由电子波函数穿过样品及物镜的相干演化所支配,该物理过程可自然映射至量子计算的门模型。我们提出QuScope,一个开源Python框架,其将完整的TEM图像形成流程表达为量子电路。N×N的电子波函数以振幅编码形式存储于2log₂N个量子比特中,每个光学元件(包括相位光栅透射、样品切片间的菲涅耳传播及像差物镜)均通过量子傅里叶变换共轭实现为对角幺正操作。在此基础上,QuScope v0.2.0实现了经验证的成像流程,涵盖相位物体近似下的常规TEM、适用于厚样品的完整多层切片CTEM与STEM。本文报道的所有量子结果均来自经典硬件上对电路的精确无噪声态矢量模拟,且每个结果均通过经典孪生实现进行验证。量子与经典多层切片出射波的保真度为1,所有物理常数均经标准参考文献验证。我们提供了每种算法的编译后量子资源估计、对决定近期硬件执行的对角合成瓶颈的分析,以及一个完全测试、文档齐全且可通过pip安装的软件包。QuScope v0.2.0可在指定URL获取。

英文摘要

Image formation in transmission electron microscopy (TEM) is governed by the coherent evolution of the electron wavefunction through the specimen and the objective lens. This physics maps naturally onto the gate model of quantum computation. We present QuScope, an open-source Python framework that expresses the complete TEM image-formation pipeline as quantum circuits. The $N\times N$ electron wavefunction is amplitude-encoded in $2\log_2 N$ qubits, and every optical element, including phase-grating transmission, Fresnel propagation between specimen slices, and the aberrated objective lens, is implemented as a diagonal unitary conjugated by quantum Fourier transforms. On this foundation, QuScope v0.2.0 implements validated imaging pipelines, covering conventional TEM under the phase-object approximation, full multislice CTEM and STEM for thick specimens. All quantum results reported here come from exact, noise-free statevector simulation of the circuits on classical hardware, and every result is validated against a classical twin implementation. Quantum and classical multislice exit waves agree to unit fidelity, and all physical constants are verified against standard references. We provide transpiled quantum-resource estimates for each algorithm, an analysis of the diagonal-synthesis bottleneck that governs near-term hardware execution, and a fully tested, documented, and pip-installable package. QuScope v0.2.0 is available at https://github.com/QuScope/QuScope.

Commentsv2 - fixes CTF plot. Open-source Python framework for quantum-circuit formulations of CTEM and STEM simulation. Code, documentation, and reproducible examples are available at https://github.com/QuScope/QuScope

论文原文

arXiv 摘要页 · PDF 原文 · HTML 原文

↑