The recognition of human actions in video streams is a challenging task in computer vision, with cardinal applications in e.g. brain-computer interface and surveillance. Deep learning has shown remarkable results recently, but can be found hard to use in practice, as its training requires large datasets and special purpose, energy-consuming hardware. In this work, we propose a scalable photonic neuro-inspired architecture based on the reservoir computing paradigm, capable of recognising video-based human actions with state-of-the-art accuracy. Our experimental optical setup comprises off-the-shelf components, and implements a large parallel recurrent neural network that is easy to train and can be scaled up to hundreds of thousands of nodes. This work paves the way towards simply reconfigurable and energy-efficient photonic information processing systems for real-time video processing.
We analyze numerically optical waveguide structures in photorefractive media induced by one or two incoherent counter-propagating (CP) Airy beams. Under nonlinear focusing conditions, we show that for a single Airy beam or for two CP beams with various input positions, multiple waveguiding structures can be photo-induced in the medium. Optical Gaussian beams can therefore be guided with a deflecting trajectory and/or even split into several output beams. These results enable new configurations for all-optical interconnections.
We propose a scalable photonic architecture for implementation of feedforward and recurrent neural networks to perform the classification of handwritten digits from the MNIST database. Our experiment exploits off-the-shelf optical and electronic components to currently achieve a network size of 16,384 nodes. Both network types are designed within the the reservoir computing paradigm with randomly weighted input and hidden layers. Using various feature extraction techniques (e.g. histograms of oriented gradients, zoning, Gabor filters) and a simple training procedure consisting of linear regression and winner-takes-all decision strategy, we demonstrate numerically and experimentally that a feedforward network allows for classification error rate of 1%, which is at the state-of-the-art for experimental implementations and remains competitive with more advanced algorithmic approaches. We also investigate recurrent networks in numerical simulations by explicitly activating the temporal dynamics, and predict a performance improvement over the feedforward configuration.
We have experimentally analyzed pattern formation in an optical system composed of a bulk photorefractive crystal subjected to a single optical feedback. In a highly nonlinear regime far above the modulational instability threshold, we are reporting on turbulent spatiotemporal dynamics that leads to rare, intense localized optical peaks. We have proven that the statistics and features of those peaks correspond to the signatures of two-dimensional spatial rogue events. These optical rogue waves occur erratically in space and time and live typically the same amount of time as the response time of the photorefractive material.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.