In-materio reservoir working at low frequencies in a Ag<sub>2</sub>S-island network

Nakajima, Motoharu; Minegishi, K.; Shimizu, Y.; Usami, Yuki; Tanaka, Hirofumi; Hasegawa, Tsuyoshi

doi:10.1039/d2nr01439d

Cited by 17 publications

(17 citation statements)

References 57 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this sense, frequency range of input signals that a reservoir accepts is an important specification. A Ag 2 S island network reservoir, used in this study, utilizes multiple physicochemical phenomena in its reservoir operation: diffusion of Ag + cations in a Ag 2 S island, growth/shrinkage of a Ag protrusion from/into a Ag 2 S island, and formation/annihilation of a Ag filament that connects adjacent Ag 2 S islands, 26) as shown in Fig. 1(b).…”

Section: Introductionmentioning

confidence: 99%

Pulse width dependent operations of a Ag₂S island network reservoir

Minegishi

Shimizu

Hasegawa

2023

Jpn. J. Appl. Phys.

Self Cite

View full text Add to dashboard Cite

The rapid growth in demand for edge artificial intelligence increases importance of physical reservoirs that work at low computational cost with low power consumption. A Ag2S island network also works as a physical reservoir, in which various physicochemical phenomena contribute to a reservoir operation. In this study, we investigated its frequency dependence and found that diffusion of Ag+ cations in a Ag2S island, which has a relaxation time of about 100 µs, plays a major role when performance is improved. MNIST classification task using an input pulse width of 100 µs resulted in the accuracy of 91 %. Iterative operations up to 10 million cycles revealed a small enough standard deviation of output, suggesting a potential for practical use of a Ag2S island network as a reservoir.

show abstract

Section: Introductionmentioning

confidence: 99%

Pulse width dependent operations of a Ag₂S island network reservoir

Minegishi

Shimizu

Hasegawa

2023

Jpn. J. Appl. Phys.

Self Cite

View full text Add to dashboard Cite

show abstract

“…Cr/Ag/BN-diffusive/Ag/Cr memristor [60] 3.5 × 10 −5 --III h) Ag 2 S/PTCDI-C8/Ag 2 S memristor [61] 8 × 10 −6 --X-99%…”

Section: Methodsmentioning

confidence: 99%

Power‐Efficient Multisensory Reservoir Computing Based on Zr‐Doped HfO₂ Memcapacitive Synapse Arrays

et al. 2023

View full text Add to dashboard Cite

Hardware implementations tailored to requirements in reservoir computing would facilitate lightweight and powerful temporal processing. Capacitive reservoirs would boost power efficiency due to their ultralow static power consumption but haven't been experimentally exploited yet. Here we report an oxide‐based memcapacitive synapse (OMC) based on Zr‐doped HfO2 (HZO) for a power‐efficient and multisensory processing reservoir computing system. The nonlinearity and state richness required for reservoir computing could be originated from the capacitively coupled polarization switching and charge trapping of hafnium oxide‐based devices. The power consumption (∼113.4 fJ/spike) and temporal processing versatility outperform most resistive reservoirs. Our system has been verified by common benchmark tasks, and it exhibits high accuracy (>94%) in recognizing multisensory information, including acoustic, electrophysiological, and mechanic modalities. As a proof‐of‐concept, a touchless user interface for virtual shopping is demonstrated based on our OMC‐based reservoir computing system, benefiting from its interference‐robust acoustic and electrophysiological perception. Our results would shed light on the development of high power‐efficient human‐machine interfaces and machine‐learning platforms.This article is protected by copyright. All rights reserved

show abstract

“…Various physical reservoirs have been used in the reservoir layer, [11][12][13][14][15][16][17] and research into materials-based reservoir computing is currently accelerating. [18][19][20] For example, reservoir computing research exists using atomic switch technologies to control the growth and contraction of Ag filaments from Ag 2 S. 12,21,22) Ag 2 S reservoirs have also been used for benchmarking tasks such as STM, PC, and MNIST image recognition and optical pattern classification. 23,24) In terms of social implementation of artificial intelligence, AI robot development using machine learning, such as deep learning, has been conducted in recent years, [25][26][27] which will meet the increasing demand for care and assistance robots.…”

Section: Introductionmentioning

confidence: 99%

Research on tactile sensation by physical reservoir computing with a robot arm and a Ag₂S reservoir

Yoshimura,

Hasegawa

2024

Jpn. J. Appl. Phys.

Self Cite

View full text Add to dashboard Cite

In recent years, physical reservoir computing has attracted much attention because of its low computational cost and low power consumption. In terms of social implementation of artificial intelligence, physical reservoir has a potential to meet the request, such as the need for AI robots to process information related to tactile sensation. It has been reported that a Ag2S polycrystalline thin film retains short-term memory and non-linearity when used as a physical reservoir. In this study, we applied the technique to tactile sensation by combining with a pressure sensor attached to a robot arm. In the object grasping task, a Ag2S physical reservoir enabled the objective recognition with the accuracy of 81.3%, although the task failed with linear regression of the direct output from the pressure sensor. We also demonstrate the potential of the system to detect anomalies in object grabbing.

show abstract

In-materio reservoir working at low frequencies in a Ag₂S-island network

Abstract: A Ag2S-island network is fabricated with surrounding electrodes to enable it to be used as a reservoir for unconventional computing. Local conductance change occurs due to the growth/shrinkage of Ag...

Cited by 17 publications

References 57 publications

Pulse width dependent operations of a Ag₂S island network reservoir

Pulse width dependent operations of a Ag₂S island network reservoir

Power‐Efficient Multisensory Reservoir Computing Based on Zr‐Doped HfO₂ Memcapacitive Synapse Arrays

Research on tactile sensation by physical reservoir computing with a robot arm and a Ag₂S reservoir

Contact Info

Product

Resources

About

In-materio reservoir working at low frequencies in a Ag2S-island network

Abstract: A Ag2S-island network is fabricated with surrounding electrodes to enable it to be used as a reservoir for unconventional computing. Local conductance change occurs due to the growth/shrinkage of Ag...

Cited by 17 publications

References 57 publications

Pulse width dependent operations of a Ag2S island network reservoir

Pulse width dependent operations of a Ag2S island network reservoir

Power‐Efficient Multisensory Reservoir Computing Based on Zr‐Doped HfO2 Memcapacitive Synapse Arrays

Research on tactile sensation by physical reservoir computing with a robot arm and a Ag2S reservoir

Contact Info

Product

Resources

About

In-materio reservoir working at low frequencies in a Ag₂S-island network

Pulse width dependent operations of a Ag₂S island network reservoir

Pulse width dependent operations of a Ag₂S island network reservoir

Power‐Efficient Multisensory Reservoir Computing Based on Zr‐Doped HfO₂ Memcapacitive Synapse Arrays

Research on tactile sensation by physical reservoir computing with a robot arm and a Ag₂S reservoir