Deep learning with coherent nanophotonic circuits
01 Jul 2017-Vol. 11, Iss: 7, pp 441-446
TL;DR: A new architecture for a fully optical neural network is demonstrated that enables a computational speed enhancement of at least two orders of magnitude and three order of magnitude in power efficiency over state-of-the-art electronics.
Abstract: Artificial Neural Networks have dramatically improved performance for many machine learning tasks. We demonstrate a new architecture for a fully optical neural network that enables a computational speed enhancement of at least two orders of magnitude and three orders of magnitude in power efficiency over state-of-the-art electronics.
Citations
More filters
01 Mar 2022
TL;DR: Using a multi-layer perceptron to equalise the residual nonlinearity from the transmission of PDM 28 Gbaud 64QAM over 400km of SSMF employing midlink optical phase conjugation, a 12-fold reduction in the BER was demonstrated in this paper .
Abstract: Using a multi-layer perceptron to equalise the residual nonlinearity from the transmission of PDM 28 Gbaud 64QAM over 400km of SSMF employing midlink optical phase conjugation, we demonstrate 12-fold reduction in the BER. © 2021 The Author(s)
4 citations
TL;DR: In this paper , a photonic implementation of a reservoir computer that exploits frequency domain multiplexing to encode neuron states is presented, and the system processes 25 comb lines simultaneously (i.e., 25 neurons) at a rate of 20 MHz.
Abstract: Reservoir computing is a brain-inspired approach for information processing, well suited to analog implementations. We report a photonic implementation of a reservoir computer that exploits frequency domain multiplexing to encode neuron states. The system processes 25 comb lines simultaneously (i.e., 25 neurons), at a rate of 20 MHz. We illustrate performances on two standard benchmark tasks: channel equalization and time series forecasting. We also demonstrate that frequency multiplexing allows output weights to be implemented in the optical domain, through optical attenuation. We discuss the perspectives for high-speed, high-performance, low-footprint implementations.
4 citations
Posted Content•
TL;DR: In this article, a single spatial phase modulator simultaneously encodes spin configurations and programs interaction strengths, which enables a promising approach for solving NP-hard problems with large-scale and ultra-fast optical computing.
Abstract: Statistical spin dynamics plays a key role to understand the working principle for novel optical Ising machines. Here we propose the gauge transformations for spatial optical spin systems, where a single spatial phase modulator simultaneously encodes spin configurations and programs interaction strengths. Thanks to gauge transformation, we experimentally evaluate the phase diagram of high-dimensional spin-glass equilibrium system with $100$ fully-connected spins. We identify that there is a spin-glass phase that the gauge-transformed spins become synchronized when disorder interactions are rather strong. Furthermore, we exploit the gauge transformation for high accuracy and good scalability in solving combinatorial optimization problems, for the spin number as large as $N=40000$. Our results show that the gauge transformation enables a promising approach for solving NP-hard problems with large-scale and ultra-fast optical computing.
4 citations
TL;DR: In this article , the role of low symmetry and broken symmetry in controlling nanoscale light and its widespread applications for optical technologies is discussed, including the physics of moiré photonics, in-plane inversion symmetry breaking for valleytronics and nonradiative state control, and parity-time symmetry breaking.
Abstract: Photonics and optoelectronics are at the foundations of widespread technologies, from high-speed internet to systems for artificial intelligence, LiDAR, and optical quantum computing. Light enables ultrafast speeds and low energy for all-optical information processing and transport, especially when confined at the nanoscale level, at which the interactions of light with matter unveil new phenomena, and the role of local symmetries becomes crucial. This Perspective discusses how symmetry violations provide unique opportunities for nanophotonics, tailoring wave interactions in nanostructures for a wide range of functionalities. In particular, we discuss broken geometrical symmetries for localized surface polaritons, the physics of moiré photonics, in-plane inversion symmetry breaking for valleytronics and nonradiative state control, time-reversal symmetry breaking for optical nonreciprocity, and parity-time symmetry breaking. Overall, our Perspective aims to present the role of low symmetry and broken symmetry in controlling nanoscale light and its widespread applications for optical technologies.
4 citations
TL;DR: In this paper , the authors investigated the effect of batch normalization layer (BatchNorm) on the noise-resistant ability of deep learning models, and showed that the presence of the BatchNorm layer negatively impacts the noise resistant property of DNNs.
Abstract: The fast execution speed and energy efficiency of analog hardware have made them a strong contender for deploying deep learning models at the edge. However, there are concerns about the presence of analog noise which causes changes to the models’ weight, leading to performance degradation of deep learning models, despite their inherent noise-resistant characteristics. The effect of the popular batch normalization layer (BatchNorm) on the noise-resistant ability of deep learning models is investigated in this work. This systematic study has been carried out by first training different models with and without the BatchNorm layer on the CIFAR10 and the CIFAR100 datasets. The weights of the resulting models are then injected with analog noise, and the performance of the models on the test dataset is obtained and compared. The results show that the presence of the BatchNorm layer negatively impacts the noise-resistant property of deep learning models, i.e., ResNet44 and VGG16 models with BatchNorm layers trained with the CIFAR10 dataset have an average normalized inference accuracy of 41.32% and 10.75% respectively compared to 91.95% and 93.80% obtained for same ResNet44 and VGG16 model without the BatchNorm layer respectively. Furthermore, the impact of the BatchNorm layer also grows with the increase of the number of BatchNorm layers.
4 citations
References
More filters
Proceedings Article•
03 Dec 2012TL;DR: The state-of-the-art performance of CNNs was achieved by Deep Convolutional Neural Networks (DCNNs) as discussed by the authors, which consists of five convolutional layers, some of which are followed by max-pooling layers, and three fully-connected layers with a final 1000-way softmax.
Abstract: We trained a large, deep convolutional neural network to classify the 1.2 million high-resolution images in the ImageNet LSVRC-2010 contest into the 1000 different classes. On the test data, we achieved top-1 and top-5 error rates of 37.5% and 17.0% which is considerably better than the previous state-of-the-art. The neural network, which has 60 million parameters and 650,000 neurons, consists of five convolutional layers, some of which are followed by max-pooling layers, and three fully-connected layers with a final 1000-way softmax. To make training faster, we used non-saturating neurons and a very efficient GPU implementation of the convolution operation. To reduce overriding in the fully-connected layers we employed a recently-developed regularization method called "dropout" that proved to be very effective. We also entered a variant of this model in the ILSVRC-2012 competition and achieved a winning top-5 test error rate of 15.3%, compared to 26.2% achieved by the second-best entry.
73,978 citations
[...]
TL;DR: Deep learning is making major advances in solving problems that have resisted the best attempts of the artificial intelligence community for many years, and will have many more successes in the near future because it requires very little engineering by hand and can easily take advantage of increases in the amount of available computation and data.
Abstract: Deep learning allows computational models that are composed of multiple processing layers to learn representations of data with multiple levels of abstraction. These methods have dramatically improved the state-of-the-art in speech recognition, visual object recognition, object detection and many other domains such as drug discovery and genomics. Deep learning discovers intricate structure in large data sets by using the backpropagation algorithm to indicate how a machine should change its internal parameters that are used to compute the representation in each layer from the representation in the previous layer. Deep convolutional nets have brought about breakthroughs in processing images, video, speech and audio, whereas recurrent nets have shone light on sequential data such as text and speech.
46,982 citations
TL;DR: This work bridges the divide between high-dimensional sensory inputs and actions, resulting in the first artificial agent that is capable of learning to excel at a diverse array of challenging tasks.
Abstract: The theory of reinforcement learning provides a normative account, deeply rooted in psychological and neuroscientific perspectives on animal behaviour, of how agents may optimize their control of an environment. To use reinforcement learning successfully in situations approaching real-world complexity, however, agents are confronted with a difficult task: they must derive efficient representations of the environment from high-dimensional sensory inputs, and use these to generalize past experience to new situations. Remarkably, humans and other animals seem to solve this problem through a harmonious combination of reinforcement learning and hierarchical sensory processing systems, the former evidenced by a wealth of neural data revealing notable parallels between the phasic signals emitted by dopaminergic neurons and temporal difference reinforcement learning algorithms. While reinforcement learning agents have achieved some successes in a variety of domains, their applicability has previously been limited to domains in which useful features can be handcrafted, or to domains with fully observed, low-dimensional state spaces. Here we use recent advances in training deep neural networks to develop a novel artificial agent, termed a deep Q-network, that can learn successful policies directly from high-dimensional sensory inputs using end-to-end reinforcement learning. We tested this agent on the challenging domain of classic Atari 2600 games. We demonstrate that the deep Q-network agent, receiving only the pixels and the game score as inputs, was able to surpass the performance of all previous algorithms and achieve a level comparable to that of a professional human games tester across a set of 49 games, using the same algorithm, network architecture and hyperparameters. This work bridges the divide between high-dimensional sensory inputs and actions, resulting in the first artificial agent that is capable of learning to excel at a diverse array of challenging tasks.
23,074 citations
"Deep learning with coherent nanopho..." refers background or methods in this paper
...The computational resolution of ONNs is limited by practical non-idealities, including (1) thermal crosstalk between phase shifters in interferometers, (2) optical coupling drift, (3) the finite precision with which an optical phase can be set (16 bits in our case), (4) photodetection noise and (5) finite photodetection dynamic range (30 dB in our case)....
[...]
...(3) Once a neural network is trained, the architecture can be passive, and computation on the optical signals will be performed without additional energy input....
[...]
...We used four instances of the OIU to realize the following matrix transformations in the spatial-mode basis: (1) U((1))Σ((1)), (2) V((1)), (3) U((2))Σ((2)) and (4) V((2))....
[...]
...Transformations (1) and (2) realize the first matrix M((1)), and (3) and (4) implement M((2))....
[...]
TL;DR: In this article, an effective way of initializing the weights that allows deep autoencoder networks to learn low-dimensional codes that work much better than principal components analysis as a tool to reduce the dimensionality of data is described.
Abstract: High-dimensional data can be converted to low-dimensional codes by training a multilayer neural network with a small central layer to reconstruct high-dimensional input vectors. Gradient descent can be used for fine-tuning the weights in such "autoencoder" networks, but this works well only if the initial weights are close to a good solution. We describe an effective way of initializing the weights that allows deep autoencoder networks to learn low-dimensional codes that work much better than principal components analysis as a tool to reduce the dimensionality of data.
16,717 citations
TL;DR: This historical survey compactly summarizes relevant work, much of it from the previous millennium, review deep supervised learning, unsupervised learning, reinforcement learning & evolutionary computation, and indirect search for short programs encoding deep and large networks.
14,635 citations
"Deep learning with coherent nanopho..." refers methods in this paper
...ANNs can be trained by feeding training data into the input layer and then computing the output by forward propagation; weighting parameters in each matrix are subsequently optimized using back propagation [16]....
[...]