A Review of Optical Neural Networks

Res-U2Net: untrained deep learning for phase retrieval and image reconstruction

Conventional deep learning-based image reconstruction methods require a large amount of training data, which can be hard to obtain in practice. Untrained deep learning methods overcome this limitation by training a network to invert a physical model of the image formation process. Here we present a novel, to our knowledge, untrained Res-U2Net model for phase retrieval. We use the extracted phase information to determine changes in an object's surface and generate a mesh representation of its 3D structure. We compare the performance of Res-U2Net phase retrieval against UNet and U2Net using images from the GDXRAY dataset.

Exploring Types of Photonic Neural Networks for Imaging and Computing-A Review

Photonic neural networks (PNNs), utilizing light-based technologies, show immense potential in artificial intelligence (AI) and computing. Compared to traditional electronic neural networks, they offer faster processing speeds, lower energy usage, and improved parallelism. Leveraging light's properties for information processing could revolutionize diverse applications, including complex calculations and advanced machine learning (ML). Furthermore, these networks could address scalability and efficiency challenges in large-scale AI systems, potentially reshaping the future of computing and AI research. In this comprehensive review, we provide current, cutting-edge insights into diverse types of PNNs crafted for both imaging and computing purposes. Additionally, we delve into the intricate challenges they encounter during implementation, while also illuminating the promising perspectives they introduce to the field.

Neuromorphic Photonics Circuits: Contemporary Review

Neuromorphic photonics is a cutting-edge fusion of neuroscience-inspired computing and photonics technology to overcome the constraints of conventional computing architectures. Its significance lies in the potential to transform information processing by mimicking the parallelism and efficiency of the human brain. Using optics and photonics principles, neuromorphic devices can execute intricate computations swiftly and with impressive energy efficiency. This innovation holds promise for advancing artificial intelligence and machine learning while addressing the limitations of traditional silicon-based computing. Neuromorphic photonics could herald a new era of computing that is more potent and draws inspiration from cognitive processes, leading to advancements in robotics, pattern recognition, and advanced data processing. This paper reviews the recent developments in neuromorphic photonic integrated circuits, applications, and current challenges.

Silicon-Based On-Chip Tunable High-Q-Factor and Low-Power Fano Resonators with Graphene Nanoheaters

Tunable and low-power microcavities are essential for large-scale photonic integrated circuits. Thermal tuning, a convenient and stable tuning method, has been widely adopted in optical neural networks and quantum information processing. Recently, graphene thermal tuning has been demonstrated to be a power-efficient technique, as it does not require thick spacers to prevent light absorption. In this paper, a silicon-based on-chip Fano resonator with graphene nanoheaters is proposed and fabricated. This novel Fano structure is achieved by introducing a scattering block, and it can be easily fabricated in large quantities. Experimental results demonstrate that the resonator has the characteristics of a high quality factor (∼31,000) and low state-switching power (∼1 mW). The temporal responses of the microcavity exhibit qualified modulation speed with 9.8 μs rise time and 16.6 μs fall time. The thermal imaging and Raman spectroscopy of graphene at different biases were also measured to intuitively show that the tuning is derived from the joule heating effect of graphene. This work provides an alternative for future large-scale tunable and low-power-consumption optical networks, and has potential applications in optical filters and switches.

Translation-invariant optical neural network for image classification

The classification performance of all-optical Convolutional Neural Networks (CNNs) is greatly influenced by components' misalignment and translation of input images in the practical applications. In this paper, we propose a free-space all-optical CNN (named Trans-ONN) which accurately classifies translated images in the horizontal, vertical, or diagonal directions. Trans-ONN takes advantages of an optical motion pooling layer which provides the translation invariance property by implementing different optical masks in the Fourier plane for classifying translated test images. Moreover, to enhance the translation invariance property, global average pooling (GAP) is utilized in the Trans-ONN structure, rather than fully connected layers. The comparative studies confirm that taking advantage of vertical and horizontal masks along GAP operation provide the best translation invariance property, compared to the alternative network models, for classifying horizontally and vertically shifted test images up to 50 pixel shifts of Kaggle Cats and Dogs, CIFAR-10, and MNIST datasets, respectively. Also, adopting the diagonal mask along GAP operation achieves the best classification accuracy for classifying translated test images in the diagonal direction for large number of pixel shifts (i.e. more than 30 pixel shifts). It is worth mentioning that the proposed translation invariant networks are capable of classifying the translated test images not included in the training procedure.

Magneto-optical diffractive deep neural network

We propose a magneto-optical diffractive deep neural network (MO-D²NN). We simulated several MO-D²NNs, each of which consists of five hidden layers made of a magnetic material that contains 100 × 100 magnetic domains with a domain width of 1 µm and an interlayer distance of 0.7 mm. The networks demonstrate a classification accuracy of > 90% for the MNIST dataset when light intensity is used as the classification measure. Moreover, an accuracy of > 80% is obtained even for a small Faraday rotation angle of π/100 rad when the angle of polarization is used as the classification measure. The MO-D²NN allows the hidden layers to be rewritten, which is not possible with previous implementations of D²NNs.