论文标题
同时多参数估计的量子生成的对抗学习
Quantum generative adversarial learning for simultaneous multiparameter estimation
论文作者
论文摘要
生成的对手学习目前是人工智能中最多产的领域之一，因为它在各种具有挑战性的任务中的表现出色，例如逼真的图像和视频产生。尽管出现了一种量子版本的对抗性学习，该学习承诺与其经典同行具有指数优势，但其实验实现和使用可访问的量子技术的潜在应用程序仍然很少探索。在这里，我们报告了基于随机梯度下降算法的自适应反馈的帮助，对量子生成对抗性学习进行了实验证明。通过将此技术应用于量子动力学的自适应表征和多个阶段的同时估计来探索其性能。这些结果表明，即使在存在有害噪声的情况下，量子生成对抗性学习的有趣优势，并为量子增强的信息处理应用铺平了道路。
Generative adversarial learning is currently one of the most prolific fields in artificial intelligence due to its great performance in a variety of challenging tasks such as photorealistic image and video generation. While a quantum version of generative adversarial learning has emerged that promises exponential advantages over its classical counterpart, its experimental implementation and potential applications with accessible quantum technologies remain explored little. Here, we report an experimental demonstration of quantum generative adversarial learning with the assistance of adaptive feedback that is based on stochastic gradient descent algorithm. Its performance is explored by applying this technique to the adaptive characterization of quantum dynamics and simultaneous estimation of multiple phases. These results indicate the intriguing advantages of quantum generative adversarial learning even in the presence of deleterious noise, and pave the way towards quantum-enhanced information processing applications.
