Quantum-noise-limited optical neural networks operating at a few quanta per activation

• URL: http://arxiv.org/abs/2307.15712v1
• Date: Fri, 28 Jul 2023 17:59:46 GMT
• Title: Quantum-noise-limited optical neural networks operating at a few quanta per activation
• Authors: Shi-Yuan Ma, Tianyu Wang, J\'er\'emie Laydevant, Logan G. Wright and Peter L. McMahon
• Abstract summary: We show that it is possible to train optical neural networks to perform deterministic image-classification tasks with high accuracy. We experimentally demonstrated MNIST classification with a test accuracy of 98% using an optical neural network with a hidden layer operating in the single-photon regime.
• Score: 5.494796517705931
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Analog physical neural networks, which hold promise for improved energy efficiency and speed compared to digital electronic neural networks, are nevertheless typically operated in a relatively high-power regime so that the signal-to-noise ratio (SNR) is large (>10). What happens if an analog system is instead operated in an ultra-low-power regime, in which the behavior of the system becomes highly stochastic and the noise is no longer a small perturbation on the signal? In this paper, we study this question in the setting of optical neural networks operated in the limit where some layers use only a single photon to cause a neuron activation. Neuron activations in this limit are dominated by quantum noise from the fundamentally probabilistic nature of single-photon detection of weak optical signals. We show that it is possible to train stochastic optical neural networks to perform deterministic image-classification tasks with high accuracy in spite of the extremely high noise (SNR ~ 1) by using a training procedure that directly models the stochastic behavior of photodetection. We experimentally demonstrated MNIST classification with a test accuracy of 98% using an optical neural network with a hidden layer operating in the single-photon regime; the optical energy used to perform the classification corresponds to 0.008 photons per multiply-accumulate (MAC) operation, which is equivalent to 0.003 attojoules of optical energy per MAC. Our experiment used >40x fewer photons per inference than previous state-of-the-art low-optical-energy demonstrations, to achieve the same accuracy of >90%. Our work shows that some extremely stochastic analog systems, including those operating in the limit where quantum noise dominates, can nevertheless be used as layers in neural networks that deterministically perform classification tasks with high accuracy if they are appropriately trained.

Related papers

• A spiking photonic neural network of 40.000 neurons, trained with rank-order coding for leveraging sparsity [0.0]
  We present a photonic neural network (SNN) comprising 40,000 neurons using off-the-shelf components. The network achieves 83.5% accuracy on MNIST using 22% of neurons, and 77.5% with 8.5% neuron utilization. This demonstration integrates photonic nonlinearity, excitability, and sparse computation, paving the way for efficient large-scale photonic neuromorphic systems.
  arXiv  Detail & Related papers  (2024-11-28T15:28:30Z)
• Optical training of large-scale Transformers and deep neural networks with direct feedback alignment [48.90869997343841]
  We experimentally implement a versatile and scalable training algorithm, called direct feedback alignment, on a hybrid electronic-photonic platform. An optical processing unit performs large-scale random matrix multiplications, which is the central operation of this algorithm, at speeds up to 1500 TeraOps. We study the compute scaling of our hybrid optical approach, and demonstrate a potential advantage for ultra-deep and wide neural networks.
  arXiv  Detail & Related papers  (2024-09-01T12:48:47Z)
• Expressivity of Neural Networks with Random Weights and Learned Biases [44.02417750529102]
  Recent work has pushed the bounds of universal approximation by showing that arbitrary functions can similarly be learned by tuning smaller subsets of parameters. We provide theoretical and numerical evidence demonstrating that feedforward neural networks with fixed random weights can be trained to perform multiple tasks by learning biases only. Our results are relevant to neuroscience, where they demonstrate the potential for behaviourally relevant changes in dynamics without modifying synaptic weights.
  arXiv  Detail & Related papers  (2024-07-01T04:25:49Z)
• Single Neuromorphic Memristor closely Emulates Multiple Synaptic Mechanisms for Energy Efficient Neural Networks [71.79257685917058]
  We demonstrate memristive nano-devices based on SrTiO3 that inherently emulate all these synaptic functions. These memristors operate in a non-filamentary, low conductance regime, which enables stable and energy efficient operation.
  arXiv  Detail & Related papers  (2024-02-26T15:01:54Z)
• All-optical modulation with single-photons using electron avalanche [69.65384453064829]
  We demonstrate all-optical modulation using a beam with single-photon intensity. Our approach opens up the possibility of terahertz-speed optical switching at the single-photon level.
  arXiv  Detail & Related papers  (2023-12-18T20:14:15Z)
• Spatially Varying Nanophotonic Neural Networks [39.1303097259564]
  Photonic processors that execute operations using photons instead of electrons promise to enable optical neural networks with ultra-low latency and power consumption. Existing optical neural networks, limited by the underlying network designs, have achieved image recognition accuracy far below that of state-of-the-art electronic neural networks.
  arXiv  Detail & Related papers  (2023-08-07T08:48:46Z)
• Globally Optimal Training of Neural Networks with Threshold Activation Functions [63.03759813952481]
  We study weight decay regularized training problems of deep neural networks with threshold activations. We derive a simplified convex optimization formulation when the dataset can be shattered at a certain layer of the network.
  arXiv  Detail & Related papers  (2023-03-06T18:59:13Z)
• Single-Shot Optical Neural Network [55.41644538483948]
  'Weight-stationary' analog optical and electronic hardware has been proposed to reduce the compute resources required by deep neural networks. We present a scalable, single-shot-per-layer weight-stationary optical processor.
  arXiv  Detail & Related papers  (2022-05-18T17:49:49Z)
• Resonant tunnelling diode nano-optoelectronic spiking nodes for neuromorphic information processing [0.0]
  We introduce an optoelectronic artificial neuron capable of operating at ultrafast rates and with low energy consumption. The proposed system combines an excitable tunnelling diode (RTD) element with a nanoscale light source.
  arXiv  Detail & Related papers  (2021-07-14T14:11:04Z)
• An optical neural network using less than 1 photon per multiplication [4.003843776219224]
  We experimentally demonstrate an optical neural network achieving 99% accuracy on handwritten-digit classification. This performance was achieved using a custom free-space optical processor. Our results provide a proof-of-principle for low-optical-power operation.
  arXiv  Detail & Related papers  (2021-04-27T20:43:23Z)
• All-optical neural network quantum state tomography [0.39146761527401414]
  We build an integrated all-optical setup for neural network QST, based on an all-optical neural network (AONN) Experiment results demonstrate the validity and efficiency of the all-optical setup. Our all-optical setup of integrated AONN-QST may shed light on replenishing the all-optical quantum network with the last brick.
  arXiv  Detail & Related papers  (2021-03-11T04:43:34Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.