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
TL;DR: In this paper, the power splitter was realized by adiabatically tapered rib waveguides with 60-nm shallow etches, which exhibited an excess loss as low as 0.06 dB at a 1550-nm wavelength and a broad operating wavelength range from 1470 nm to 1570 nm.
Abstract: A silicon photonic 3-dB power splitter is one of the essential components to demonstrate large-scale silicon photonic integrated circuits (PICs), and can be utilized to implement modulators, 1 × 2 switches, and 1 × N power splitters for various PIC applications. In this paper, we reported the design and experimental demonstration of low-loss and broadband silicon photonic 3-dB power splitters. The power splitter was realized by adiabatically tapered rib waveguides with 60-nm shallow etches. The shallow-etched rib waveguides offered strong coupling and relaxed critical dimensions (a taper tip width of 200 nm and gap spacing of 300 nm). The fabricated device exhibited an excess loss as low as 0.06 dB at a 1550-nm wavelength and a broad operating wavelength range from 1470 nm to 1570 nm. The relaxed critical dimensions (≥200 nm) make the power splitter compatible with standard fabrication processes of existing silicon photonics foundries.
8 citations
02 Sep 2021
TL;DR: An optical neural-network architecture for edge computing that takes advantage of wavelength multiplexing, high-bandwidth modulation, and integration detection to allow large-scale neural networks to be run on low-power edge devices accessible by an optical link is introduced.
Abstract: We introduce an optical neural-network architecture for edge computing that takes advantage of wavelength multiplexing, high-bandwidth modulation, and integration detection. Our protocol consists of a server and a client, which divide the task of neural-network inference into two steps: (1) a difficult step of optical weight distribution, performed at the server and (2) an easy step of modulation and integration detection, performed at the edge device. This arrangement allows for large-scale neural networks to be run on low-power edge devices accessible by an optical link. We perform simulations to estimate the speed and energy limits of this scheme.
8 citations
13 Jul 2020
TL;DR: In this paper, an on-chip programmable multi-level nonvolatile photonic memory is used as node in a photonic neural network that effortlessly performs inference at the edge of the network as a passive and reprogrammable filter.
Abstract: Here we demonstrate an on-chip programmable multi-level non-volatile photonic memory used as node in a photonic neural network that effortlessly perform inference at the edge of the network as a passive and reprogrammable filter.
8 citations
TL;DR: In this paper , a comprehensive summary of various optical logic gate schemes including spatial encoding of light field, semiconductor optical amplifiers (SOA), highly nonlinear fiber (HNLF), microscale and nanoscale waveguides, and photonic crystal structures is presented.
Abstract: Optical computing and optical neural network have gained increasing attention in recent years because of their potential advantages of parallel processing at the speed of light and low power consumption by comparison with electronic computing. The optical implementation of the fundamental building blocks of a digital computer, i.e. logic gates, has been investigated extensively in the past few decades. Optical logic gate computing is an alternative approach to various analogue optical computing architectures. In this paper, the latest development of optical logic gate computing with different kinds of implementations is reviewed. Firstly, the basic concepts of analogue and digital computing with logic gates in the electronic and optical domains are introduced. And then a comprehensive summary of various optical logic gate schemes including spatial encoding of light field, semiconductor optical amplifiers (SOA), highly nonlinear fiber (HNLF), microscale and nanoscale waveguides, and photonic crystal structures is presented. To conclude, the formidable challenges in developing practical all-optical logic gates are analyzed and the prospects of the future are discussed. al. All-optical logic gate computing for high-speed parallel information processing. Opto-Electron Sci 1 , 220010 (2022).
8 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]....
[...]