-
Emergent self-adaptation in an integrated photonic neural network for backpropagation-free learning
Authors:
Alessio Lugnan,
Samarth Aggarwal,
Frank Brückerhoff-Plückelmann,
C. David Wright,
Wolfram H. P. Pernice,
Harish Bhaskaran,
Peter Bienstman
Abstract:
Plastic self-adaptation, nonlinear recurrent dynamics and multi-scale memory are desired features in hardware implementations of neural networks, because they enable them to learn, adapt and process information similarly to the way biological brains do. In this work, we experimentally demonstrate these properties occurring in arrays of photonic neurons. Importantly, this is realised autonomously i…
▽ More
Plastic self-adaptation, nonlinear recurrent dynamics and multi-scale memory are desired features in hardware implementations of neural networks, because they enable them to learn, adapt and process information similarly to the way biological brains do. In this work, we experimentally demonstrate these properties occurring in arrays of photonic neurons. Importantly, this is realised autonomously in an emergent fashion, without the need for an external controller setting weights and without explicit feedback of a global reward signal. Using a hierarchy of such arrays coupled to a backpropagation-free training algorithm based on simple logistic regression, we are able to achieve a performance of 98.2% on the MNIST task, a popular benchmark task looking at classification of written digits. The plastic nodes consist of silicon photonics microring resonators covered by a patch of phase-change material that implements nonvolatile memory. The system is compact, robust, and straightforward to scale up through the use of multiple wavelengths. Moreover, it constitutes a unique platform to test and efficiently implement biologically plausible learning schemes at a high processing speed.
△ Less
Submitted 6 December, 2023;
originally announced December 2023.
-
Photonic neural networks based on integrated silicon microresonators
Authors:
Stefano Biasi,
Giovanni Donati,
Alessio Lugnan,
Mattia Mancinelli,
Emiliano Staffoli,
Lorenzo Pavesi
Abstract:
The recent progress of artificial intelligence (AI) has boosted the computational possibilities in fields where standard computers are not able to perform. The AI paradigm is to emulate human intelligence and therefore breaks the familiar architecture on which digital computers are based. In particular, neuromorphic computing, artificial neural networks (ANN) and deep learning models mimic how the…
▽ More
The recent progress of artificial intelligence (AI) has boosted the computational possibilities in fields where standard computers are not able to perform. The AI paradigm is to emulate human intelligence and therefore breaks the familiar architecture on which digital computers are based. In particular, neuromorphic computing, artificial neural networks (ANN) and deep learning models mimic how the brain computes. Large networks of interconnected neurons whose synapsis are individually strengthened or weakened during the learning phase find many applications. With this respect, photonics is a suitable platform to implement ANN hardware thanks to its speed, low power dissipation and multi-wavelength opportunities. One photonic device candidate to perform as an optical neuron is the optical microring resonator. Indeed microring resonators show both a nonlinear response and a capability of optical energy storage, which can be interpreted as a fading memory. Moreover, by using silicon photonics, the photonic integrated circuits can be fabricated in volume and with integrated electronics on board. For these reasons, here, we describe the physics of silicon microring resonators and of arrays of microring resonators for application in neuromorphic computing. We describe different types of ANNs from feed-forward networks to photonics extreme learning machines and reservoir computing. In addition, we discuss also hybrid systems where silicon microresonators are coupled to other active materials. this review aims to introduce the basics and to discuss the most recent developments in the field.
△ Less
Submitted 7 June, 2023;
originally announced June 2023.
-
Training a spiking neural network on an event-based label-free flow cytometry dataset
Authors:
Muhammed Gouda,
Steven Abreu,
Alessio Lugnan,
Peter Bienstman
Abstract:
Imaging flow cytometry systems aim to analyze a huge number of cells or micro-particles based on their physical characteristics. The vast majority of current systems acquire a large amount of images which are used to train deep artificial neural networks. However, this approach increases both the latency and power consumption of the final apparatus. In this work-in-progress, we combine an event-ba…
▽ More
Imaging flow cytometry systems aim to analyze a huge number of cells or micro-particles based on their physical characteristics. The vast majority of current systems acquire a large amount of images which are used to train deep artificial neural networks. However, this approach increases both the latency and power consumption of the final apparatus. In this work-in-progress, we combine an event-based camera with a free-space optical setup to obtain spikes for each particle passing in a microfluidic channel. A spiking neural network is trained on the collected dataset, resulting in 97.7% mean training accuracy and 93.5% mean testing accuracy for the fully event-based classification pipeline.
△ Less
Submitted 19 March, 2023;
originally announced March 2023.