RESUMEN
A spiking neural network (SNN) is a type of artificial neural network that operates based on discrete spikes to process timing information, similar to the manner in which the human brain processes real-world problems. In this paper, we propose a new spiking neural network (SNN) based on conventional, biologically plausible paradigms, such as the leaky integrate-and-fire model, spike timing-dependent plasticity, and the adaptive spiking threshold, by suggesting new biological models; that is, dynamic inhibition weight change, a synaptic wiring method, and Bayesian inference. The proposed network is designed for image recognition tasks, which are frequently used to evaluate the performance of conventional deep neural networks. To manifest the bio-realistic neural architecture, the learning is unsupervised, and the inhibition weight is dynamically changed; this, in turn, affects the synaptic wiring method based on Hebbian learning and the neuronal population. In the inference phase, Bayesian inference successfully classifies the input digits by counting the spikes from the responding neurons. The experimental results demonstrate that the proposed biological model ensures a performance improvement compared with other biologically plausible SNN models.
Asunto(s)
Aprendizaje , Neuronas , Humanos , Teorema de Bayes , Encéfalo , Redes Neurales de la ComputaciónRESUMEN
Inspired by information processing in biological systems, sensor-combined edge-computing systems attract attention requesting artificial sensory neurons as essential ingredients. Here, we introduce a simple and versatile structure of artificial sensory neurons based on a novel three-terminal Ovonic threshold switch (3T-OTS), which features an electrically controllable threshold voltage (Vth). Combined with a sensor driving an output voltage, this 3T-OTS generates spikes with a frequency depending on an external stimulus. As a proof of concept, we have built an artificial retinal ganglion cell (RGC) by combining a 3T-OTS and a photodiode. Furthermore, this artificial RGC is combined with the reservoir-computing technique to perform a classification of chest X-ray images for normal, viral pneumonia, and COVID-19 infections, releasing the recognition accuracy of about 86.5%. These results indicate that the 3T-OTS is highly promising for applications in neuromorphic sensory systems, providing a building block for energy-efficient in-sensor computing devices.
Asunto(s)
COVID-19 , Redes Neurales de la Computación , Humanos , SARS-CoV-2 , Células Receptoras SensorialesRESUMEN
The diamond-graphite hybrid thin film with low-dimensional nanostructure (e.g., nitrogen-included ultrananocrystalline diamond (N-UNCD) or the alike), has been employed in many impactful breakthrough applications. However, the detailed picture behind the bottom-up evolution of such intriguing carbon nanostructure is far from clarified yet. Here, the authors clarify it, through the concerted efforts of microscopic, physical, and electrochemical analyses for a series of samples synthesized by hot-filament chemical vapor deposition using methane-hydrogen precursor gas, based on the hydrogen-dependent surface reconstruction of nanodiamond and on the substrate-temperature-dependent variation of the growth species (atomic hydrogen and methyl radical) concentration near substrate. The clarified picture provides insights for a drastic enhancement in the electrochemical activities of the hybrid thin film, concerning the detection of important biomolecule, that is, ascorbic acid, uric acid, and dopamine: their limits of detections are 490, 35, and 25 nm, respectively, which are among the best of the all-carbon thin film electrodes in the literature. This work also enables a simple and effective way of strongly enhancing AA detection.
Asunto(s)
Grafito , Nanoestructuras , Diamante/química , Dopamina/análisis , Técnicas Electroquímicas , Electrodos , Grafito/química , Nanoestructuras/químicaRESUMEN
Associative multimodal artificial intelligence (AMAI) has gained significant attention across various fields, yet its implementation poses challenges due to the burden on computing and memory resources. To address these challenges, researchers have paid increasing attention to neuromorphic devices based on novel materials and structures, which can implement classical conditioning behaviors with simplified circuitry. Herein, we introduce an artificial multimodal neuron device that shows not only the acquisition behavior but also the extinction and the spontaneous recovery behaviors for the first time. Being composed of an ovonic threshold switch (OTS)-based neuron device, a conductive bridge memristor (CBM)-based synapse device, and a few passive electrical elements, such observed behaviors of this neuron device are explained in terms of the electroforming and the diffusion of metallic ions in the CBM. We believe that the proposed associative learning neuron device will shed light on the way of developing large-scale AMAI systems by providing inspiration to devise an associative learning network with improved energy efficiency.
RESUMEN
As there is an increasing need for an efficient solver of combinatorial optimization problems, much interest is paid to the Ising machine, which is a novel physics-driven computing system composed of coupled oscillators mimicking the dynamics of the system of coupled electronic spins. In this work, we propose an energy-efficient nano-oscillator, called OTSNO, which is composed of an Ovonic Threshold Switch (OTS) and an electrical resistor. We demonstrate that the OTSNO shows the synchronization behavior, an essential property for the realization of an Ising machine. Furthermore, we have discovered that the capacitive coupling is advantageous over the resistive coupling for the hardware implementation of an Ising solver by providing a larger margin of the variations of components. Finally, we implement an Ising machine composed of capacitively-coupled OTSNOs to demonstrate that the solution to a 14-node MaxCut problem can be obtained in 40 µs while consuming no more than 2.3 µJ of energy. Compared to a previous hardware implementation of the phase-transition nano-oscillator (PTNO)-based Ising machine, the OTSNO-based Ising machine in this work shows the performance of the increased speed by more than one order while consuming less energy by about an order.
RESUMEN
Neuromorphic computing research is being actively pursued to address the challenges posed by the need for energy-efficient processing of big data. One of the promising approaches to tackle the challenges is the hardware implementation of spiking neural networks (SNNs) with bio-plausible learning rules. Numerous research works have been done to implement the SNN hardware with different synaptic plasticity rules to emulate human brain operations. While a standard spike-timing-dependent-plasticity (STDP) rule is emulated in many SNN hardware, the various STDP rules found in the biological brain have rarely been implemented in hardware. This study proposes a CMOS-memristor hybrid synapse circuit for the hardware implementation of a Ca ion-based plasticity model to emulate the various STDP curves. The memristor was adopted as a memory device in the CMOS synapse circuit because memristors have been identified as promising candidates for analog non-volatile memory devices in terms of energy efficiency and scalability. The circuit design was divided into four sub-blocks based on biological behavior, exploiting the non-volatile and analog state properties of memristors. The circuit was designed to vary weights using an H-bridge circuit structure and PWM modulation. The various STDP curves have been emulated in one CMOS-memristor hybrid circuit, and furthermore a simple neural network operation was demonstrated for associative learning such as Pavlovian conditioning. The proposed circuit is expected to facilitate large-scale operations for neuromorphic computing through its scale-up.
RESUMEN
Neuromorphic cognitive computing offers a bio-inspired means to approach the natural intelligence of biological neural systems in silicon integrated circuits. Typically, such circuits either reproduce biophysical neuronal dynamics in great detail as tools for computational neuroscience, or abstract away the biology by simplifying the functional forms of neural computation in large-scale systems for machine intelligence with high integration density and energy efficiency. Here we report a hybrid which offers biophysical realism in the emulation of multi-compartmental neuronal network dynamics at very large scale with high implementation efficiency, and yet with high flexibility in configuring the functional form and the network topology. The integrate-and-fire array transceiver (IFAT) chip emulates the continuous-time analog membrane dynamics of 65 k two-compartment neurons with conductance-based synapses. Fired action potentials are registered as address-event encoded output spikes, while the four types of synapses coupling to each neuron are activated by address-event decoded input spikes for fully reconfigurable synaptic connectivity, facilitating virtual wiring as implemented by routing address-event spikes externally through synaptic routing table. Peak conductance strength of synapse activation specified by the address-event input spans three decades of dynamic range, digitally controlled by pulse width and amplitude modulation (PWAM) of the drive voltage activating the log-domain linear synapse circuit. Two nested levels of micro-pipelining in the IFAT architecture improve both throughput and efficiency of synaptic input. This two-tier micro-pipelining results in a measured sustained peak throughput of 73 Mspikes/s and overall chip-level energy efficiency of 22 pJ/spike. Non-uniformity in digitally encoded synapse strength due to analog mismatch is mitigated through single-point digital offset calibration. Combined with the flexibly layered and recurrent synaptic connectivity provided by hierarchical address-event routing of registered spike events through external memory, the IFAT lends itself to efficient large-scale emulation of general biophysical spiking neural networks, as well as rate-based mapping of rectified linear unit (ReLU) neural activations.
RESUMEN
Brain-like artificial intelligence in electronics can be built efficiently by understanding the connectivity of neuronal circuitry. The concept of neural connectivity inference with a two-dimensional cross-bar structure memristor array is indicated in recent studies; however, large-scale implementation is challenging owing to device variations and the requirement of online parameter adaptation. This study proposes a neural connectivity inference method with one-dimensional spiking neurons using spike timing-dependent plasticity and presynaptic spike-driven spike timing-dependent plasticity learning rules, designed for a large-scale neuromorphic system. The proposed learning process decreases the number of spiking neurons by half. We simulate 12 ground-truth neural networks comprising one-dimensional eight and 64 neurons. We analyze the correlation between the neural connectivity of the ground truth and spiking neural networks using the Matthews correlation coefficient. In addition, we analyze the sensitivity and specificity of inference. Validation using the presynaptic spike-driven spike timing-dependent plasticity learning rule implies a potential approach for large-scale neural network inference with real hardware realization of large-scale neuromorphic systems.
Asunto(s)
Inteligencia Artificial , Plasticidad Neuronal , Potenciales de Acción/fisiología , Plasticidad Neuronal/fisiología , Redes Neurales de la Computación , Neuronas/fisiologíaRESUMEN
The electrosynthesis of formate from CO2 can mitigate environmental issues while providing an economically valuable product. Although stannic oxide is a good catalytic material for formate production, a metallic phase is formed under high reduction overpotentials, reducing its activity. Here, using a fluorine-doped tin oxide catalyst, a high Faradaic efficiency for formate (95% at 100 mA cm-2) and a maximum partial current density of 330 mA cm-2 (at 400 mA cm-2) is achieved for the electroreduction of CO2. Furthermore, the formate selectivity (≈90%) is nearly constant over 7 days of operation at a current density of 100 mA cm-2. In-situ/operando spectroscopies reveal that the fluorine dopant plays a critical role in maintaining the high oxidation state of Sn, leading to enhanced durability at high current densities. First-principle calculation also suggests that the fluorine-doped tin oxide surface could provide a thermodynamically stable environment to form HCOO* intermediate than tin oxide surface. These findings suggest a simple and efficient approach for designing active and durable electrocatalysts for the electrosynthesis of formate from CO2.
RESUMEN
Memristors, or memristive devices, have attracted tremendous interest in neuromorphic hardware implementation. However, the high electric-field dependence in conventional filamentary memristors results in either digital-like conductance updates or gradual switching only in a limited dynamic range. Here, we address the switching parameter, the reduction probability of Ag cations in the switching medium, and ultimately demonstrate a cluster-type analogue memristor. Ti nanoclusters are embedded into densified amorphous Si for the following reasons: low standard reduction potential, thermodynamic miscibility with Si, and alloy formation with Ag. These Ti clusters effectively induce the electrochemical reduction activity of Ag cations and allow linear potentiation/depression in tandem with a large conductance range (~244) and long data retention (~99% at 1 hour). Moreover, according to the reduction potentials of incorporated metals (Pt, Ta, W, and Ti), the extent of linearity improvement is selectively tuneable. Image processing simulation proves that the Ti4.8%:a-Si device can fully function with high accuracy as an ideal synaptic model.
Asunto(s)
Ingeniería , Metales , Aleaciones , Simulación por Computador , Oxidación-ReducciónRESUMEN
Mid-infrared wavelengths are called the molecular fingerprint region because it contains the fundamental vibrational modes inherent to the substances of interest. Since the mid-infrared spectrum can provide non-destructive identification and quantitative analysis of unknown substances, miniaturized mid-infrared spectrometers for on-site diagnosis have attained great concern. Filter-array based on-chip spectrometer has been regarded as a promising alternative. In this study, we explore a way of applying a pillar-type plasmonic nanodiscs array, which is advantageous not only for excellent tunability of resonance wavelength but also for 2-dimensional integration through a single layer process, to the multispectral filter array for the on-chip spectrometer. We theoretically and experimentally investigated the optical properties of multi-periodic triangular lattices of metal nanodiscs array that act as stopband filters in the mid-infrared region. Soft-mold reverse nanoimprint lithography with a subsequent lift-off process was employed to fabricate the multispectral filter array and its filter function was successfully extracted using a Fourier transform infrared microscope. With the measured filter function, we tested the feasibility of target spectrum reconstruction using a Tikhonov regularization method for an ill-posed linear problem and evaluated its applicability to the infrared spectroscopic sensor that monitors an oil condition. These results not only verify that the multispectral filter array composed of stopband filters based on the metal nanodiscs array when combined with the spectrum reconstruction technique, has great potential for use to a miniaturized mid-infrared on-chip spectrometer, but also provide effective guidance for the filter design.
RESUMEN
Among many artificial neural networks, the research on Spike Neural Network (SNN), which mimics the energy-efficient signal system in the brain, is drawing much attention. Memristor is a promising candidate as a synaptic component for hardware implementation of SNN, but several non-ideal device properties are making it challengeable. In this work, we conducted an SNN simulation by adding a device model with a non-linear weight update to test the impact on SNN performance. We found that SNN has a strong tolerance for the device non-linearity and the network can keep the accuracy high if a device meets one of the two conditions: 1. symmetric LTP and LTD curves and 2. positive non-linearity factors for both LTP and LTD. The reason was analyzed in terms of the balance between network parameters as well as the variability of weight. The results are considered to be a piece of useful prior information for the future implementation of emerging device-based neuromorphic hardware.
RESUMEN
A novel nano-plasmonic sensing platform based on vertical conductive bridge was suggested as an alternative geometry for taking full advantages of unique properties of conductive junction while substantially alleviating burdens in lithographic process. The effects of various geometrical parameters on the plasmonic properties were systematically investigated. Theoretical simulation on this structure demonstrates that the presence of vertical conductive bridge with smaller diameter sandwiched between two adjacent thin nanodiscs excites a bridged mode very similar to the charge transfer plasmon and exhibits a remarkable enhancement in the extinction efficiency and the sensitivity when the electric field of incident light is parallel to the conductive bridge. Furthermore, for the electric field perpendicular to the bridge, another interesting feature is observed that two magnetic resonance modes are excited symmetrically through open-gaps on both sides of the bridge together with strongly enhanced electric field intensity, which provides a very favorable environment as a surface enhanced Raman scattering substrate for fluid analysis. These results verify a great potential and versatility of our approach for use as a nanoplasmonic sensing platform. In addition, we demonstrated the feasibility of fabrication process of vertical conductive bridge and high tunability in controlling the bridge width.
RESUMEN
We report the complementary resistive switching (CRS) behaviors in a tantalum-oxide based resistive switching memory device that reversibly changes its switching mode between bipolar switching (BRS) and CRS in a single memory cell depending on the operation (compliance current) and fabrication (oxygen scavenger layer thickness) conditions. In addition, the origin of the switching mode transition was investigated through electrical and optical measurement, where the conductance is believed to be determined by two factors: formation of conductive filament and modulation of Schottky barrier. This result helps design a resistive switching device with desirable and stable switching behavior.
RESUMEN
Neuromorphic computing is of great interest among researchers interested in overcoming the von Neumann computing bottleneck. A synaptic device, one of the key components to realize a neuromorphic system, has a weight that indicates the strength of the connection between two neurons, and updating this weight must have linear and symmetric characteristics. Especially, a transistor-type device has a gate terminal, separating the processes of reading and updating the conductivity, used as a synaptic weight to prevent sneak path current issues during synaptic operations. In this study, we fabricate a top-gated flash memory device based on two-dimensional (2D) materials, MoS2 and graphene, as a channel and a floating gate, respectively, and Al2O3 and HfO2 to increase the tunneling efficiency. We demonstrate the linear weight updates and repeatable characteristics of applying negative/positive pulses, and also emulate spike timing-dependent plasticity (STDP), one of the learning rules in a spiking neural network (SNN).
RESUMEN
A multichannel neural interface system is an important tool for various types of neuroscientific studies. For the electrical interface with a biological system, high-precision high-speed data recording and various types of stimulation capability are required. In addition, real-time signal processing is an important feature in the implementation of a real-time closed-loop system without unwanted substantial delay for feedback stimulation. Online spike sorting, the process of assigning neural spikes to an identified group of neurons or clusters, is a necessary step to make a closed-loop path in real time, but massive memory-space requirements commonly limit hardware implementations. Here, we present a 128-channel field-programmable gate array (FPGA)-based real-time closed-loop bidirectional neural interface system. The system supports 128 channels for simultaneous signal recording and eight selectable channels for stimulation. A modular 64-channel analog front-end (AFE) provides scalability and a parameterized specification of the AFE supports the recording of various electrophysiological signal types with 1.59 ± 0.76 root-mean-square noise. The stimulator supports both voltage-controlled and current-controlled arbitrarily shaped waveforms with the programmable amplitude and duration of pulse. An empirical algorithm for online real-time spike sorting is implemented in an FPGA. The spike-sorting is performed by template matching, and templates are created by an online real-time unsupervised learning process. A memory saving technique, called dynamic cache organizing, is proposed to reduce the memory requirement down to 6 kbit per channel and modular implementation improves the scalability for further extensions.
Asunto(s)
Prótesis Neurales , Algoritmos , Animales , Corteza Cerebral/citología , Sistemas de Computación , Retroalimentación , Femenino , Humanos , Aprendizaje Automático , Microelectrodos , Neuronas/fisiología , Embarazo , Cultivo Primario de Células , Ratas , Reproducibilidad de los Resultados , Procesamiento de Señales Asistido por Computador , Relación Señal-RuidoRESUMEN
We present a hierarchical address-event routing (HiAER) architecture for scalable communication of neural and synaptic spike events between neuromorphic processors, implemented with five Xilinx Spartan-6 field-programmable gate arrays and four custom analog neuromophic integrated circuits serving 262k neurons and 262M synapses. The architecture extends the single-bus address-event representation protocol to a hierarchy of multiple nested buses, routing events across increasing scales of spatial distance. The HiAER protocol provides individually programmable axonal delay in addition to strength for each synapse, lending itself toward biologically plausible neural network architectures, and scales across a range of hierarchies suitable for multichip and multiboard systems in reconfigurable large-scale neuromorphic systems. We show approximately linear scaling of net global synaptic event throughput with number of routing nodes in the network, at 3.6×107 synaptic events per second per 16k-neuron node in the hierarchy.
RESUMEN
We covalently bound fluoropolymer (FP) films by plasma treatment followed by thermal pressing at temperatures below their melting point and fabricated an adhesion-metal-free flexible gold electrode array entirely encapsulated by the FP film, excepting the active electrode openings. The fabricated device was chemically resistant and was modified to have a lower impedance and efficient charge injection capability. The fabricated device was evaluated in vivo in rats and was confirmed to record the epidural epileptiform activity induced by chemical administration. The chemically inert nature of FPs and the gold electrode is expected to facilitate reliable neural interfacing without abiotic issues. Plasma treatment-induced covalent binding of FP films can also be utilized in a variety of applications requiring durability, such as implantable biosensors and sensor platforms operating under chemically harsh environments, including humid conditions.
RESUMEN
Synchrony and temporal coding in the central nervous system, as the source of local field potentials and complex neural dynamics, arises from precise timing relationships between spike action population events across neuronal assemblies. Recently it has been shown that coincidence detection based on spike event timing also presents a robust neural code invariant to additive incoherent noise from desynchronized and unrelated inputs. We present spike-based coincidence detection using integrate-and-fire neural membrane dynamics along with pooled conductance-based synaptic dynamics in a hierarchical address-event architecture. Within this architecture, we encode each synaptic event with parameters that govern synaptic connectivity, synaptic strength, and axonal delay with additional global configurable parameters that govern neural and synaptic temporal dynamics. Spike-based coincidence detection is observed and analyzed in measurements on a log-domain analog VLSI implementation of the integrate-and-fire neuron and conductance-based synapse dynamics.