The sensitivity of optical measurement is ultimately constrained by the shot noise to the standard quantum limit.It has become a common concept that beating this limit requires quantum resources.A deep-learning neural network free of quantum principle has the capability of removing classical noise from images, but it is unclear in reducing quantum noise.In a coincidence-imaging experiment, we show that quantum-resource-free deep learning can be exploited to surpass the standard quantum limit via the photon-number-dependent nonlinear feedback during training.Using an effective classical light with photon flux of about 9 × 10 4 photons per second, our deep-learning-based scheme achieves a 14 dB improvement in signal-to-noise ratio with respect to the standard quantum limit.
The sensitivity of optical measurement is ultimately constrained by the shot noise to the standard quantum limit.It has become a common concept that beating this limit requires quantum resources.A deep-learning neural network free of quantum principle has the capability of removing classical noise from images, but it is unclear in reducing quantum noise.In a coincidence-imaging experiment, we show that quantum-resource-free deep learning can be exploited to surpass the standard quantum limit via the photon-number-dependent nonlinear feedback during training.Using an effective classical light with photon flux of about 9 × 10 4 photons per second, our deep-learning-based scheme achieves a 14 dB improvement in signal-to-noise ratio with respect to the standard quantum limit.
Deep neural networks have been shown to achieve exceptional performance for computer vision tasks like image recognition, segmentation, and reconstruction or denoising. Here, we evaluate the ultimate performance limits of deep convolutional neural network models for image reconstruction, by comparing them against the standard quantum limit set by shot-noise and the Heisenberg limit on precision. We train U-Net models on images of natural objects illuminated with coherent states of light, and find that the average mean-squared error of the reconstructions can surpass the standard quantum limit, and in some cases reaches the Heisenberg limit. Further, we train models on well-parameterized images for which we can calculate the quantum Cramér-Rao bound to determine the minimum possible measurable variance of an estimated parameter for a given probe state. We find the mean-squared error of the model predictions reaches these bounds calculated for the parameters, across a variety of parameterized images. These results suggest that deep convolutional neural networks can learn to become the optimal estimators allowed by the laws of physics, performing parameter estimation and image reconstruction at the ultimate possible limits of precision for the case of classical illumination of the object.
Quantum imaging using photon pairs with strong quantum correlations has been harnessed to bring quantum advantages to various fields from biological imaging to range finding. Such inherent non-classical properties support the extraction of more valid signals to build photon-limited images, even in low-light conditions where the shot noise becomes dominant as light decreases to a single-photon level. Numerical optimization algorithms are possible but require thousands of photon-sparse frames, and they are thus unavailable in real time. We demonstrate fast correlated-photon imaging enhanced by deep learning as an intelligent computational strategy to discover a deeper structure in big data. Our work verifies that a convolutional neural network can efficiently solve inverse imaging problems associated with strong shot noise and background noise (electronic noise, scattered light). Our results show that we can overcome limitations due to the trade-off between imaging speed and image quality by pushing the low-light imaging technique to the single-photon level in real time, which enables deep-learning-enhanced quantum imaging for real-life applications.
Correlated photon pairs, carrying strong quantum correlations, have been harnessed to bring quantum advantages to various fields from biological imaging to range finding. Such inherent non-classical properties support extracting more valid signals to build photon-limited images even in low flux-level, where the shot noise becomes dominant as light source decreases to single-photon level. Optimization by numerical reconstruction algorithms is possible but require thousands of photon-sparse frames, thus unavailable in real time. Here, we present an experimental fast correlated-photon imaging enhanced by deep learning, showing an intelligent computational strategy to discover deeper structure in big data. Convolutional neural network is found being able to efficiently solve image inverse problems associated with strong shot noise and background noise (electronic noise, scattered light). Our results fill the key gap in incompatibility between imaging speed and image quality by pushing low-light imaging technique to the regime of real-time and single-photon level, opening up an avenue to deep leaning-enhanced quantum imaging for real-life applications.
Abstract The phase sensitivity of photonic NOON states scales O(1/N), which reaches the Heisenberg limit and indicates a great potential in high-quality optical phase sensing. However, the NOON states with large photon number N are experimentally difficult both to prepare and to operate. Such a fact severely limits their practical use. In this article, we soften the requirements for high-quality imaging based on NOON states with large N by introducing deep-learning methods. Specifically, we show that, with the help of deep-learning network, the fluctuation of the images obtained by the NOON states when N = 2 can be reduced to that of the currently infeasible imaging by the NOON states when N = 8. We numerically investigate our results obtained by two types of deep-learning models—deep neural network and convolutional denoising autoencoders, and characterize the imaging quality using the root mean square error. By comparison, we find that small-N NOON state imaging data is sufficient for training the deep-learning models of our schemes, which supports its direct application to the imaging processes.
Quantum-inspired imaging techniques have been proven to be effective for LiDAR with the advances of single photon detectors and computational algorithms. However, due to the disturbance of background noise and the varies of signal in outdoor environment, the performance of LiDAR is still far from its ultimate limit set by the quantum fluctuations of coherent probe light. In this work, we propose and demonstrate a LiDAR from the detection perspective for approaching the standard quantum-limited performance. The photon numbers of echo signals are recorded by a photon-number-resolving detector and applied to overcome heavy background noise through an active photon number filter in the LiDAR. It can approach the standard quantum limit in intensity estimation in a wide photon-flux range, and achieve a Fisher information of only 0.04 dB less than the quantum Fisher information when the mean signal photon number is 10. Experimentally, a noise-free target reconstruction and imaging is demonstrated in the daytime by the proposed LiDAR. It also performs better in reflectivity resolution when taking only 1/1000 of the measurements based on on/off detection. This work provides a fundamental strategy for constructing a LiDAR to quickly extract targets and identify materials in complex environments, which is important for intelligent agents such as autonomous vehicles.
Recently, several proof of principle experiments have demonstrated the advantages of quantum technologies over classical schemes. The present challenge is to surpass the limits of proof of principle demonstrations to approach real applications. This letter presents such an achievement in the field of quantum enhanced imaging. In particular, we describe the realization of a sub-shot-noise wide field microscope based on spatially multi-mode non-classical photon number correlations in twin beams. The microscope produces realtime images of 8000 pixels at full resolution, for a 500 μm2 field of view, with noise reduced to 80% of the shot noise level (for each pixel), which is suitable for absorption imaging of complex structures. By fast post-elaboration, specifically applying a quantum enhanced median filter, the noise can be further reduced (to <30% of the shot noise level) by setting a trade-off with the resolution, thus achieving the best sensitivity per incident photon reported in absorption microscopy. A new microscope instantly captures thousands of photons from faint objects with signal-to-noise ratios exceeding classical limits. Imaging of sensitive objects, such as live cells, limits the optical power that lasers can apply. Now, Alice Meda from Italy's INRIM and colleagues have developed a microscope that ’squeezes‘ light to overcome intensity-limiting shot noise caused by quantum fluctuations of photons. Their sub-shot-noise microscope uses a nonlinear crystal to generate two high correlated light beams impossible for classical waves. Microscale pixels recorded intensities when one beam strikes a sample can have their shot noise reduced by 20% by subtracting correlated patterns from the second reference beam. This quantum-enhanced noise reduction enables the microscope to operate for high sensitive, dynamic, wide field, Q imaging applications.