All-Optical Image Identification with Programmable Matrix Transformation
arXiv:2104.02474 · doi:10.1364/OE.430281
Abstract
An optical neural network is proposed and demonstrated with programmable matrix transformation and nonlinear activation function of photodetection (square-law detection). Based on discrete phase-coherent spatial modes, the dimensionality of programmable optical matrix operations is 30~37, which is implemented by spatial light modulators. With this architecture, all-optical classification tasks of handwritten digits, objects and depth images are performed on the same platform with high accuracy. Due to the parallel nature of matrix multiplication, the processing speed of our proposed architecture is potentially as high as7.4T~74T FLOPs per second (with 10~100GHz detector)
References in corpus (9)
- Deep Learning with Coherent Nanophotonic Circuits
- All-Optical Machine Learning Using Diffractive Deep Neural Networks
- Photonics for artificial intelligence and neuromorphic computing
- 11 TeraFLOPs per second photonic convolutional accelerator for deep learning optical neural networks
- All Optical Neural Network with Nonlinear Activation Functions
- Reprogrammable Electro-Optic Nonlinear Activation Functions for Optical Neural Networks
- Class-specific Differential Detection in Diffractive Optical Neural Networks Improves Inference Accuracy
- Universal linear optical operations on discrete phase-coherent spatial modes
- Programmable coherent linear quantum operations with high-dimensional optical spatial modes