-
On-Chip Optical Convolutional Neural Networks
Authors:
Hengameh Bagherian,
Scott Skirlo,
Yichen Shen,
Huaiyu Meng,
Vladimir Ceperic,
Marin Soljacic
Abstract:
Convolutional Neural Networks (CNNs) are a class of Artificial Neural Networks(ANNs) that employ the method of convolving input images with filter-kernels for object recognition and classification purposes. In this paper, we propose a photonics circuit architecture which could consume a fraction of energy per inference compared with state of the art electronics.
Convolutional Neural Networks (CNNs) are a class of Artificial Neural Networks(ANNs) that employ the method of convolving input images with filter-kernels for object recognition and classification purposes. In this paper, we propose a photonics circuit architecture which could consume a fraction of energy per inference compared with state of the art electronics.
△ Less
Submitted 16 August, 2018; v1 submitted 9 August, 2018;
originally announced August 2018.
-
Tunable Efficient Unitary Neural Networks (EUNN) and their application to RNNs
Authors:
Li **g,
Yichen Shen,
Tena Dubček,
John Peurifoy,
Scott Skirlo,
Yann LeCun,
Max Tegmark,
Marin Soljačić
Abstract:
Using unitary (instead of general) matrices in artificial neural networks (ANNs) is a promising way to solve the gradient explosion/vanishing problem, as well as to enable ANNs to learn long-term correlations in the data. This approach appears particularly promising for Recurrent Neural Networks (RNNs). In this work, we present a new architecture for implementing an Efficient Unitary Neural Networ…
▽ More
Using unitary (instead of general) matrices in artificial neural networks (ANNs) is a promising way to solve the gradient explosion/vanishing problem, as well as to enable ANNs to learn long-term correlations in the data. This approach appears particularly promising for Recurrent Neural Networks (RNNs). In this work, we present a new architecture for implementing an Efficient Unitary Neural Network (EUNNs); its main advantages can be summarized as follows. Firstly, the representation capacity of the unitary space in an EUNN is fully tunable, ranging from a subspace of SU(N) to the entire unitary space. Secondly, the computational complexity for training an EUNN is merely $\mathcal{O}(1)$ per parameter. Finally, we test the performance of EUNNs on the standard copying task, the pixel-permuted MNIST digit recognition benchmark as well as the Speech Prediction Test (TIMIT). We find that our architecture significantly outperforms both other state-of-the-art unitary RNNs and the LSTM architecture, in terms of the final performance and/or the wall-clock training speed. EUNNs are thus promising alternatives to RNNs and LSTMs for a wide variety of applications.
△ Less
Submitted 3 April, 2017; v1 submitted 15 December, 2016;
originally announced December 2016.
-
Deep Learning with Coherent Nanophotonic Circuits
Authors:
Yichen Shen,
Nicholas C. Harris,
Scott Skirlo,
Mihika Prabhu,
Tom Baehr-Jones,
Michael Hochberg,
Xin Sun,
Shijie Zhao,
Hugo Larochelle,
Dirk Englund,
Marin Soljacic
Abstract:
Artificial Neural Networks are computational network models inspired by signal processing in the brain. These models have dramatically improved the performance of many learning tasks, including speech and object recognition. However, today's computing hardware is inefficient at implementing neural networks, in large part because much of it was designed for von Neumann computing schemes. Significan…
▽ More
Artificial Neural Networks are computational network models inspired by signal processing in the brain. These models have dramatically improved the performance of many learning tasks, including speech and object recognition. However, today's computing hardware is inefficient at implementing neural networks, in large part because much of it was designed for von Neumann computing schemes. Significant effort has been made to develop electronic architectures tuned to implement artificial neural networks that improve upon both computational speed and energy efficiency. Here, we propose a new architecture for a fully-optical neural network that, using unique advantages of optics, promises a computational speed enhancement of at least two orders of magnitude over the state-of-the-art and three orders of magnitude in power efficiency for conventional learning tasks. We experimentally demonstrate essential parts of our architecture using a programmable nanophotonic processor.
△ Less
Submitted 7 October, 2016;
originally announced October 2016.
-
Millimeter Wave Absorber for Secure Identification
Authors:
Scott A. Skirlo,
Jonathan T. Richard,
Magued Nasr,
Martin S. Heimbeck,
John D. Joannopoulos,
Marin Soljacic,
Henry O. Everitt,
Lawrence Domash
Abstract:
We demonstrate thin, flexible, metamaterial films with a strong, narrowband, polarization- and angle-insensitive absorption designed for wavelengths near one millimeter. These structures, fabricated by photolithography on a commercially available, copper-backed polyimide substrate, are nearly indistinguishable to the unaided human eye but can be easily observed by imaging at the resonance frequenc…
▽ More
We demonstrate thin, flexible, metamaterial films with a strong, narrowband, polarization- and angle-insensitive absorption designed for wavelengths near one millimeter. These structures, fabricated by photolithography on a commercially available, copper-backed polyimide substrate, are nearly indistinguishable to the unaided human eye but can be easily observed by imaging at the resonance frequency of the film. We demonstrate that these patterns can be used to mark or barcode objects for secure identification with a terahertz imaging system.
△ Less
Submitted 1 June, 2016;
originally announced June 2016.
-
Experimental Observation of Large Chern numbers in Photonic Crystals
Authors:
Scott A. Skirlo,
Ling Lu,
Qinghui Yan,
Yuichi Igarashi,
John Joannopoulos,
Marin Soljacic
Abstract:
Despite great interest in the quantum anomalous Hall phase and its analogs, all experimental studies in electronic and bosonic systems have been limited to a Chern number of one. Here, we perform microwave transmission measurements in the bulk and at the edge of ferrimagnetic photonic crystals. Bandgaps with large Chern numbers of 2, 3, and 4 are present in the experimental results which show exce…
▽ More
Despite great interest in the quantum anomalous Hall phase and its analogs, all experimental studies in electronic and bosonic systems have been limited to a Chern number of one. Here, we perform microwave transmission measurements in the bulk and at the edge of ferrimagnetic photonic crystals. Bandgaps with large Chern numbers of 2, 3, and 4 are present in the experimental results which show excellent agreement with theory. We measure the mode profiles and Fourier transform them to produce dispersion relations of the edge modes, whose number and direction match our Chern number calculations.
△ Less
Submitted 1 December, 2015; v1 submitted 16 April, 2015;
originally announced April 2015.
-
Quantum Čerenkov Radiation: Spectral Cutoffs and the Role of Spin and Orbital Angular Momentum
Authors:
Ido Kaminer,
Maor Mutzafi,
Amir Levy,
Gal Harari,
Hanan Herzig Sheinfux,
Scott Skirlo,
Jonathan Nemirovsky,
John D. Joannopoulos,
Mordechai Segev,
Marin Soljačić
Abstract:
We show that the well-known Čerenkov Effect contains new phenomena arising from the quantum nature of charged particles. The Čerenkov transition amplitudes allow coupling between the charged particle and the emitted photon through their orbital angular momentum (OAM) and spin, by scattering into preferred angles and polarizations. Importantly, the spectral response reveals a discontinuity immediat…
▽ More
We show that the well-known Čerenkov Effect contains new phenomena arising from the quantum nature of charged particles. The Čerenkov transition amplitudes allow coupling between the charged particle and the emitted photon through their orbital angular momentum (OAM) and spin, by scattering into preferred angles and polarizations. Importantly, the spectral response reveals a discontinuity immediately below a frequency cutoff that can occur in the optical region. Specifically, with proper sha** of electron beams (ebeams), we predict that the traditional Čerenkov radiation angle splits into two distinctive cones of photonic shockwaves. One of the shockwaves can move along a backward cone, otherwise considered impossible for Čerenkov radiation in ordinary matter. Our findings are observable for ebeams with realistic parameters, offering new applications including novel quantum optics sources, and open a new realm for Čerenkov detectors involving the spin and orbital angular momentum of charged particles.
△ Less
Submitted 13 September, 2015; v1 submitted 1 November, 2014;
originally announced November 2014.
-
Binary matrices of optimal autocorrelations as alignment marks
Authors:
Scott A. Skirlo,
Ling Lu,
Marin Soljačić
Abstract:
We define a new class of binary matrices by maximizing the peak-sidelobe distances in the aperiodic autocorrelations. These matrices can be used as robust position marks for in-plane spatial alignment. The optimal square matrices of dimensions up to 7 by 7 and optimal diagonally-symmetric matrices of 8 by 8 and 9 by 9 were found by exhaustive searches.
We define a new class of binary matrices by maximizing the peak-sidelobe distances in the aperiodic autocorrelations. These matrices can be used as robust position marks for in-plane spatial alignment. The optimal square matrices of dimensions up to 7 by 7 and optimal diagonally-symmetric matrices of 8 by 8 and 9 by 9 were found by exhaustive searches.
△ Less
Submitted 28 August, 2014;
originally announced August 2014.
