An all-optical neuron with sigmoid activation function

Mourgias-Alexandris, George; Tsakyridis, Apostolos; Passalis, Nikolaos; Tefas, Anastasios; Vyrsokinos, K.; Pleros, Nikos

doi:10.1364/oe.27.009620

Cited by 174 publications

(65 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…where l x is the neuron values and l b is the bias at layer l, and f is the nonlinear activation function. There are many activation functions, the most prominent of which are sigmoid [15], tanh [16] and Relu [17].…”

Section: Deep Learning Modelsmentioning

confidence: 99%

Physics-Aware Deep Learning on Multiphase Flow Problems

Lin¹

2021

View full text Add to dashboard Cite

Section: Deep Learning Modelsmentioning

confidence: 99%

Physics-Aware Deep Learning on Multiphase Flow Problems

Lin¹

2021

View full text Add to dashboard Cite

“…Later, in [44], experimental implementation with passive nodes is provided and compared with numerical results with SOA-based neuron and sigmoid functions, showing similar performance for spoken digit classification. A significant effort has also been dedicated to mimic typical sigmoid nonlinear functions using photonics circuits, e.g., using SOAs [45] and quantum-dot lasers [46]. A comprehensive overview of the most significant contributions can be found in [19], [47], [48].…”

Section: B Rc For Signal Equalization: State-of-the-artmentioning

confidence: 99%

Reservoir-Computing Based Equalization With Optical Pre-Processing for Short-Reach Optical Transmission

Ros

Ranzini

Bülow

et al. 2020

IEEE J. Select. Topics Quantum Electron.

View full text Add to dashboard Cite

“…The weighted signals are then combined into a WDM signal and sent to the nonlinear function to provide a single wavelength neuron output. The nonlinear activation function can be realized in several ways, e.g., by employing the combination of a photodetector and a modulator [26], saturable absorbers [25], excitable lasers [28,29], wavelength converters [30], and phase change materials [31]. In this simulation work, we use a photodetector and off-line processing for nonlinear function and we mainly focus on the operation of weighted addition for the matrix multiplication.…”

Section: Photonic Deep Neural Network With Weight-soasmentioning

confidence: 99%

Numerical Simulation of an InP Photonic Integrated Cross-Connect for Deep Neural Networks on Chip

Shi

Calabretta

2020

Applied Sciences

View full text Add to dashboard Cite

We propose a novel photonic accelerator architecture based on a broadcast-and-weight approach for a deep neural network through a photonic integrated cross-connect. The single neuron and the complete neural network operation are numerically simulated. The weight calibration and weighted addition are reproduced and demonstrated to behave as in the experimental measurements. A dynamic range higher than 25 dB is predicted, in line with the measurements. The weighted addition operation is also simulated and analyzed as a function of the optical crosstalk and the number of input colors involved. In particular, while an increase in optical crosstalk negatively influences the simulated error, a greater number of channels results in better performance. The iris flower classification problem is solved by implementing the weight matrix of a trained three-layer deep neural network. The performance of the corresponding photonic implementation is numerically investigated by tuning the optical crosstalk and waveguide loss, in order to anticipate energy consumption per operation. The analysis of the prediction error as a function of the optical crosstalk per layer suggests that the first layer is essential to the final accuracy. The ultimate accuracy shows a quasi-linear dependence between the prediction accuracy and the errors per layer for a normalized root mean square error lower than 0.09, suggesting that there is a maximum level of error permitted at the first layer for guaranteeing a final accuracy higher than 89%. However, it is still possible to find good local minima even for an error higher than 0.09, due to the stochastic nature of the network we are analyzing. Lower levels of path losses allow for half the power consumption at the matrix multiplication unit, for the same error level, offering opportunities for further improved performance. The good agreement between the simulations and the experiments offers a solid base for studying the scalability of this kind of network.

show abstract

An all-optical neuron with sigmoid activation function

Cited by 174 publications

References 22 publications

Physics-Aware Deep Learning on Multiphase Flow Problems

Physics-Aware Deep Learning on Multiphase Flow Problems

Reservoir-Computing Based Equalization With Optical Pre-Processing for Short-Reach Optical Transmission

Numerical Simulation of an InP Photonic Integrated Cross-Connect for Deep Neural Networks on Chip

Contact Info

Product

Resources

About