Researchers Develop Optical Neural Network with Quantum-Like Speedup

Researchers developed a new type of optical neural network that exhibits a "quantum speedup" in training.

It uses classical optical correlations as information carriers to mimic quantum computing.

The network has convolutional and pooling operations on correlated optical states.

It shows faster convergence when learning some datasets and can identify quantum state properties.

This method could enable algorithmically enhanced optical neural networks to benefit big data processing.