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 authors presented a free-space optical ANN with diffraction-based linear weight summation and nonlinear activation enabled by the saturable absorption of thermal atoms, and demonstrated that image classification of handwritten digits using only a single layer and observed 6 percent improvement in classification accuracy compared to a linear model.
Abstract: As artificial neural networks (ANNs) continue to make strides in wide-ranging and diverse fields of technology, the search for more efficient hardware implementations beyond conventional electronics is gaining traction. In particular, optical implementations potentially offer extraordinary gains in terms of speed and reduced energy consumption due to intrinsic parallelism of free-space optics. At the same time, a physical nonlinearity, a crucial ingredient of an ANN, is not easy to realize in free-space optics, which restricts the potential of this platform. This problem is further exacerbated by the need to perform the nonlinear activation also in parallel for each data point to preserve the benefit of linear free-space optics. Here, we present a free-space optical ANN with diffraction-based linear weight summation and nonlinear activation enabled by the saturable absorption of thermal atoms. We demonstrate, via both simulation and experiment, image classification of handwritten digits using only a single layer and observed 6-percent improvement in classification accuracy due to the optical nonlinearity compared to a linear model. Our platform preserves the massive parallelism of free-space optics even with physical nonlinearity, and thus opens the way for novel designs and wider deployment of optical ANNs.
25 citations
TL;DR: This work mainly focuses on the mathematical operations that are decomposed from theoretical models of ANNs and their corresponding optical implementations; these include matrix multiplication, nonlinear activation, convolution, and learning algorithms realized via optical approaches.
Abstract: Conventional neural networks are software simulations of artificial neural networks (ANNs) implemented on von Neumann machines. This technology has recently encountered bottlenecks in terms of computing speed and energy consumption, leading to increased research interest in optical neural networks (ONNs), which are expected to become the basis for the next generation of artificial intelligence. To provide a better understanding of ONNs and to motivate further developments in this field, previous studies of ONN are reviewed in this article. Our work mainly focuses on the mathematical operations that are decomposed from theoretical models of ANNs and their corresponding optical implementations; these include matrix multiplication, nonlinear activation, convolution, and learning algorithms realized via optical approaches. Some fundamental information about ANNs is also introduced to make this work friendlier to non-experts.
25 citations
TL;DR: A rationale for a neuromorphic photonics processor is provided, envisioning its architecture and a compiler, and how it can be interfaced with a general purpose computer, i.e. a CPU, as a coprocessor to target specific applications.
Abstract: Abstract Microelectronic computers have encountered challenges in meeting all of today’s demands for information processing. Meeting these demands will require the development of unconventional computers employing alternative processing models and new device physics. Neural network models have come to dominate modern machine learning algorithms, and specialized electronic hardware has been developed to implement them more efficiently. A silicon photonic integration industry promises to bring manufacturing ecosystems normally reserved for microelectronics to photonics. Photonic devices have already found simple analog signal processing niches where electronics cannot provide sufficient bandwidth and reconfigurability. In order to solve more complex information processing problems, they will have to adopt a processing model that generalizes and scales. Neuromorphic photonics aims to map physical models of optoelectronic systems to abstract models of neural networks. It represents a new opportunity for machine information processing on sub-nanosecond timescales, with application to mathematical programming, intelligent radio frequency signal processing, and real-time control. The strategy of neuromorphic engineering is to externalize the risk of developing computational theory alongside hardware. The strategy of remaining compatible with silicon photonics externalizes the risk of platform development. In this perspective article, we provide a rationale for a neuromorphic photonics processor, envisioning its architecture and a compiler. We also discuss how it can be interfaced with a general purpose computer, i.e. a CPU, as a coprocessor to target specific applications. This paper is intended for a wide audience and provides a roadmap for expanding research in the direction of transforming neuromorphic photonics into a viable and useful candidate for accelerating neuromorphic computing.
25 citations
TL;DR: This work reports an artificial neural pathway (ANP) based on a memristor synapse to emulate neuromorphic learning behaviors and provides a promising strategy toward the design of intelligent autonomous devices and bioinspired robots through Memristor-based systems.
Abstract: Animals execute intelligent and efficient interactions with their surroundings through neural pathways, exhibiting learning, memory, and cognition. Artificial autonomous devices that generate self-optimizing feedback mimicking biological systems are essential in pursuing future intelligent robots. Here, we report an artificial neural pathway (ANP) based on a memristor synapse to emulate neuromorphic learning behaviors. In our ANP, optical stimulations are detected and converted into electrical signals through a flexible perovskite photoreceptor. The acquired electrical signals are further processed in a zeolitic imidazolate frameworks-8 (ZIF-8)-based memristor device. By controlling the growth of the ZIF-8 nanoparticles, the conductance of the memristor can be finely modulated with electrical stimulations to mimic the modulation of synaptic plasticity. The device is employed in the ANP to implement synaptic functions of learning and memory. Subsequently, the synaptic feedbacks are used to direct a robotic arm to perform responding motions. Upon repeatedly "reviewing" the optical stimulation, the ANP is able to learn, memorize, and complete the specific motions. This work provides a promising strategy toward the design of intelligent autonomous devices and bioinspired robots through memristor-based systems.
25 citations
TL;DR: The findings highlight that the present method would provide new opportunities for creating low loss WG lattices at large depth, and establish strong dependence of Li on the WG size offers a unique route to improve WG performance.
Abstract: We propose a strategy of temperature gradient assisted femtosecond laser writing for elaboration of low loss waveguides (WGs) over a large depth in glass. The matter flow driven by the temperature distribution is responsible for forming a highly densified WG core with tunable size. Importantly, the unique position of the guiding core outside the focus allows for abating the influence of laser energy redistribution and inscribing low loss deep WGs. A low insertion loss (Li) of 0.6 dB at 1550 nm is achieved for WGs at the depth from 300 µm to 900 µm. Establishing strong dependence of Li on the WG size offers a unique route to improve WG performance. These findings highlight that the present method would provide new opportunities for creating low loss WG lattices at large depth.
25 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]....
[...]