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: A novel adaptive initialization method that employs auxiliary tasks to estimate the optimal initialization variance for each layer of a network and can be directly used with any photonic activation function, without further requiring any other kind of fine-tuning.
9 citations
TL;DR: These results offer a versatile method of light management in nanostructures with prospects to improve the performance of optoelectronic devices including nanoscale LEDs, nanolasers, single photon sources, photodetectors, and solar cells.
Abstract: This work reports on high extraction efficiency in subwavelength GaAs/AlGaAs semiconductor nanopillars. We achieve up to 37-fold enhancement of the photoluminescence (PL) intensity from sub-micrometer (sub-µm) pillars without requiring back reflectors, high-Q dielectric cavities, nor large 2D arrays or plasmonic effects. This is a result of a large extraction efficiency for nanopillars <500 nm width, estimated in the range of 33-57%, which is much larger than the typical low efficiency (∼2%) of micrometer pillars limited by total internal reflection. Time-resolved PL measurements allow us to estimate the nonradiative surface recombination of fabricated pillars. We conclusively show that vertical-emitting nanopillar-based LEDs, in the best case scenario of both reduced surface recombination and efficient light out-coupling, have the potential to achieve notable large external quantum efficiency (∼45%), whereas the efficiency of large µm-pillar planar LEDs, without further methods, saturates at ∼2%. These results offer a versatile method of light management in nanostructures with prospects to improve the performance of optoelectronic devices including nanoscale LEDs, nanolasers, single photon sources, photodetectors, and solar cells.
9 citations
TL;DR: In this article, the angular momentum (AM)-involved information optics and its implementation in nanophotonic device is discussed. And a perspective in the combination of this emerging field with the optical artificial intelligence has been given.
Abstract: Light has played a crucial role in the age of information technology and has facilitated the soaring development of information optics. The ever-increasing demand for high-capacity optical devices has prompted the use of physically-orthogonal dimensions of light for optical multiplexing. Recent advances in nanotechnology, mainly stemming from functionalized nanomaterials and powerful nanofabrication tools, have propelled the fusion of optical multiplexing and nanophotonics (the study of light at a nanoscale and of its interactions with nanostructures) by enabling ultrahigh-capacity information technology. This review aims to introduce the emerging concept of angular momentum (AM)-involved information optics and its implementation in nanophotonic device. Firstly, the previous researches on the manipulation of spin angular momentum (SAM) and orbital angulat momentum (OAM) by nanostructures have been reviewed. We then summarize the SAM multiplexing technology on the platform of metasurface. Particularly, we elaborately summarized our recent progresses in the area of information optics, including OAM holography and on-chip AM multiplexing technology. Finally, a perspective in the combination of this emerging field with the optical artificial intelligence has been given.
9 citations
TL;DR: In this article, the authors review heterogeneously integrated optical modulators and switches for the next-generation silicon photonic platform, which is expected to enable high modulation efficiency, low loss, low power consumption, small device footprint, etc.
Abstract: The realization of a silicon optical phase shifter marked a cornerstone for the development of silicon photonics, and it is expected that optical interconnects based on the technology relax the explosive datacom growth in data centers. High-performance silicon optical modulators and switches, integrated into a chip, play a very important role in optical transceivers, encoding electrical signals onto the light at high speed and routing the optical signals, respectively. The development of the devices is continuously required to meet the ever-increasing data traffic at higher performance and lower cost. Therefore, heterogeneous integration is one of the highly promising approaches, expected to enable high modulation efficiency, low loss, low power consumption, small device footprint, etc. Therefore, we review heterogeneously integrated optical modulators and switches for the next-generation silicon photonic platform.
9 citations
08 Mar 2020
TL;DR: A monolithically integrated SOA-based photonic neuron is demonstrated, including both the weighted addition and a wavelength converter with tunable laser as nonlinear function, allowing for lossless computation of 8 Giga operation/s with an 89% accuracy.
Abstract: We demonstrate a monolithically integrated SOA-based photonic neuron, including both the weighted addition and a wavelength converter with tunable laser as nonlinear function, allowing for lossless computation of 8 Giga operation/s with an 89% accuracy.
9 citations
Cites background from "Deep learning with coherent nanopho..."
...Recently, photonic integrated approaches have been proposed to realize the linear [3-5] and nonlinear [6] functions, but so far these implementations have relied on hybrid integration schemes [3], external input lasers [6] and the involvement of power-consuming O/E/O conversions [4], hindering the realization of a scalable photonic neural network....
[...]
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]....
[...]