Wave-based extreme deep learning based on non-linear time-Floquet entanglement
arXiv:2107.08564 · doi:10.1038/s41467-022-30297-5
Abstract
Wave-based analog signal processing holds the promise of extremely fast, on-the-fly, power-efficient data processing, occurring as a wave propagates through an artificially engineered medium. Yet, due to the fundamentally weak non-linearities of traditional wave materials, such analog processors have been so far largely confined to simple linear projections such as image edge detection or matrix multiplications. Complex neuromorphic computing tasks, which inherently require strong non-linearities, have so far remained out-of-reach of wave-based solutions, with a few attempts that implemented non-linearities on the digital front, or used weak and inflexible non-linear sensors, restraining the learning performance. Here, we tackle this issue by demonstrating the relevance of Time-Floquet physics to induce a strong non-linear entanglement between signal inputs at different frequencies, enabling a power-efficient and versatile wave platform for analog extreme deep learning involving a single, uniformly modulated dielectric layer and a scattering medium. We prove the efficiency of the method for extreme learning machines and reservoir computing to solve a range of challenging learning tasks, from forecasting chaotic time series to the simultaneous classification of distinct datasets. Our results open the way for wave-based machine learning with high energy efficiency, speed, and scalability.
arXiv comment23 pages, 9 figures
References in corpus (32)
- Double-slit photoelectron interference in strong-field ionization of the neon dimer
- Deep Learning with Coherent Nanophotonic Circuits
- All-Optical Machine Learning Using Diffractive Deep Neural Networks
- Floquet Topological Insulators for Sound
- All Optical Neural Network with Nonlinear Activation Functions
- Reinforcement Learning in a large scale photonic Recurrent Neural Network
- Space-time gradient metasurfaces
- Large-Scale Optical Neural Networks based on Photoelectric Multiplication
- Wave Physics as an Analog Recurrent Neural Network
- Nanoscale neural network using non-linear spin-wave interference
- Mixer-Duplexer-Antenna Leaky-Wave System Based on Periodic Space-Time Modulation
- Theory of neuromorphic computing by waves: machine learning by rogue waves, dispersive shocks, and solitons
- Scalable Optical Learning Operator
- Non-Reciprocal gain in non-Hermitian time-Floquet Systems
- Serrodyne frequency translation using time-modulated metasurfaces
- Tutorial: Photonic Neural Networks in Delay Systems
- Digital Metasurface Based on Graphene: An Application to Beam Steering in Terahertz Plasmonic Antennas
- Full-Duplex Nonreciprocal-Beam-Steering Metasurfaces Comprising Time-Modulated Twin Meta-Atoms
- Random Projections through multiple optical scattering: Approximating kernels at the speed of light
- Full-duplex Reflective Beamsteering Metasurface Featuring Magnetless Nonreciprocal Amplification
- Photonic extreme learning machine by free-space optical propagation
- Generalized Optical Signal Processing Based on Multi-Operator Metasurfaces Synthesized by Susceptibility Tensors
- Turning Optical Complex Media into Universal Reconfigurable Linear Operators by Wavefront Shaping
- Electrically Tunable Harmonics in Time-modulated Metasurfaces for Wavefront Engineering
- A single inverse-designed photonic structure that performs parallel computing
- Nonreciprocal Phased-Array Antennas
- Reciprocal Metasurfaces for On-axis Reflective Optical Computing
- Parallel Optical Computing Based on MIMO Metasurface Processors with Asymmetric Optical Response
- Self-biased Tri-state Power-Multiplexed Digital Metasurface Operating at Microwave Frequencies
- Scaling up Echo-State Networks with multiple light scattering
- Switchable and Simultaneous Spatiotemporal Analog Computing
- Asymmetric Metal-dielectric Meta-cylinders and their Potential Applications from Engineering Scattering Patterns to Spatial Optical Signal Processing