Abstract:Scalable, high-capacity, and low-power computing architecture is the primary assurance for increasingly manifold and large-scale machine learning tasks. Traditional electronic artificial agents by conventional power-hungry processors have faced the issues of energy and scaling walls, hindering them from the sustainable performance improvement and iterative multi-task learning. Referring to another modality of light, photonic computing has been progressively applied in high-efficient neuromorphic systems. Here, we innovate a reconfigurable lifelong-learning optical neural network (L 2 ONN), for highly-integrated tens-of-task machine intelligence with elaborated algorithm-hardware co-design. Benefiting from the inherent sparsity and parallelism in massive photonic connections, L 2 ONN learns each single task by adaptively activating sparse photonic neuron connections in the coherent light field, while incrementally acquiring expertise on various tasks by gradually enlarging the activation. The multi-task optical features are parallelly processed by multi-spectrum representations allocated with different wavelengths. Extensive evaluations on free-space and on-chip architectures confirm that for the first time, L 2 ONN avoided the catastrophic forgetting issue of photonic computing, owning versatile skills on challenging tens-of-tasks (vision classification, voice recognition, medical diagnosis, etc.) with a single model. Particularly, L 2 ONN achieves more than an order of magnitude higher efficiency than the representative electronic artificial neural networks, and 14× larger capacity than existing optical neural networks while maintaining competitive performance on each individual task. The proposed photonic neuromorphic architecture points out a new form of lifelong learning scheme, permitting terminal/edge AI systems with light-speed efficiency and unprecedented scalability.

Fixed-Weight learning Neural Networks on Optical Hardware

Optical Analog Computing Devices Designed by Deep Neural Network

Countering Variations and Thermal Effects for Accurate Optical Neural Networks.

Fixed-weight on-line learning

Optical Implementation of Neural Learning Algorithms Based on Cross-gain Modulation in a Semiconductor Optical Amplifier

Annealing-inspired training of an optical neural network with ternary weights

Optical neural network via loose neuron array and functional learning

Forward-Forward Training of an Optical Neural Network

All-optical spiking neurosynaptic networks with self-learning capabilities

Efficient On-Chip Learning for Optical Neural Networks Through Power-Aware Sparse Zeroth-Order Optimization

An optical neural network using less than 1 photon per multiplication

A Compact Butterfly-Style Silicon Photonic–Electronic Neural Chip for Hardware-Efficient Deep Learning

Photonic neuromorphic architecture for tens-of-task lifelong learning

Genetically programmable optical random neural networks

Optical Spike-Timing-dependent Plasticity with Weight-Dependent Learning Window and Reward Modulation.

Spatially Varying Nanophotonic Neural Networks

High-throughput optical neural networks based on temporal computing

Training neural networks with end-to-end optical backpropagation

Optical training of large-scale Transformers and deep neural networks with direct feedback alignment

Deep Learning with Photonic Neural Cellular Automata

Fully forward mode training for optical neural networks