Takashi Morie

Takashi Morie

Auto-generated profile, based on publicly available information

Takashi Morie

Graduate School of Life Science and Systems Engineering, Kyushu Institute of Technology

Research Focus & Metrics

Academic Career Length

57Years (1970-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
19 July 2021 in IEEE Transactions on Neural Networks and Learning Systems
Verified
Journal Article
Closed access
YSYusuke Sakemi
Yusuke Sakemi
KMKai Morino
Kai Morino
TMTakashi Morie
Takashi Morie
KAKazuyuki Aihara
Kazuyuki Aihara
Spiking neural networks (SNNs) are brain-inspired mathematical models with the ability to process information in the form of spikes. SNNs are expected to provide not only new machine-learning algorithms but also energy-efficient computational models when implemented in very-large-scale integration (VLSI) circuits. In this article, we propose a novel supervised learning algorithm for SNNs based on temporal coding. A spiking neuron in this algorithm is designed to facilitate analog VLSI implementations with analog resistive memory, by which ultrahigh energy efficiency can be achieved. We also propose several techniques to improve the performance on recognition tasks and show that the classification accuracy of the proposed algorithm is as high as that of the state-of-the-art temporal coding SNN algorithms on the MNIST and Fashion-MNIST datasets. Finally, we discuss the robustness of the proposed SNNs against variations that arise from the device manufacturing process and are unavoidable in analog VLSI implementation. We also propose a technique to suppress the effects of variations in the manufacturing process on the recognition performance.

1 January 2020 in Chaos: An Interdisciplinary Journal of Nonlinear Science
Verified
Journal Article
Open access
PDF file
YIY. Itoh
Y. Itoh
SUS. Uenohara
S. Uenohara
MAM. Adachi
M. Adachi
TMT. Morie
T. Morie
KAK. Aihara
K. Aihara
Bifurcation-diagram reconstruction estimates various attractors of a system without observing all of them but only from observing several attractors with different parameter values. Therefore, the bifurcation-diagram reconstruction can be used to investigate how attractors change with the parameter values, especially for real-world engineering and physical systems for which only a limited number of attractors can be observed. Although bifurcation diagrams of various systems have been reconstructed from time-series data generated in numerical experiments, the systems that have been targeted for reconstructing bifurcation diagrams from time series measured from physical phenomena so far have only been continuous-time dynamical systems. In this paper, we reconstruct bifurcation diagrams only from time-series data generated by electronic circuits in discrete-time dynamical systems with different parameter values. The generated time-series datasets are perturbed by dynamical noise and contaminated by observational noise. To reconstruct the bifurcation diagrams only from the time-series datasets, we use an extreme learning machine as a time-series predictor because it has a good generalization property. Hereby, we expect that the bifurcation-diagram reconstruction with the extreme learning machine is robust against dynamical noise and observational noise. For quantitatively verifying the robustness, the Lyapunov exponents of the reconstructed bifurcation diagrams are compared with those of the bifurcation diagrams generated in numerical experiments and by the electronic circuits.

20 June 2022 in IEEE Transactions on Circuits and Systems II: Express Briefs
Verified
Journal Article
Open access
ONOsamu Nomura
Osamu Nomura
YSYusuke Sakemi
Yusuke Sakemi
THTakeo Hosomi
Takeo Hosomi
TMTakashi Morie
Takashi Morie
Spiking neural networks (SNNs) more closely mimic the human brain than artificial neural networks (ANNs). For SNNs, time-to-first-spike (TTFS) encoding, which represents the output values of neurons based on the timing of a single spike, has been proposed as a promising model to reduce power consumption. Adversarial attacks that can lead ANNs to misrecognize images have been reported in many studies. However, the characteristics of TTFS-based SNNs trained using a backpropagation algorithm against adversarial attacks have not yet been clarified. In particular, the dependence of the robustness against adversarial attacks on spike timings has not been investigated. In this brief, we investigated the robustness of SNNs against adversarial attacks and compared it with that of an ANN. We found that SNNs trained with the appropriate temporal penalty settings are more robust against adversarial images than ANNs.

28 December 2020 in IEEE Access
Verified
Journal Article
Open access
MYMasatoshi Yamaguchi
Masatoshi Yamaguchi
GIGoki Iwamoto
Goki Iwamoto
YNYuta Nishimura
Yuta Nishimura
HTHakaru TAMUKOH
Hakaru TAMUKOH
TMTakashi Morie
Takashi Morie
This paper proposes a time-domain analog calculations model based on a pulse-width modulation (PWM) approach for neural network calculations including weighted-sum or multiply-and-accumulate calculation and rectified-linear unit operation. We also propose very-large-scale integration (VLSI) circuits to implement the proposed model. Unlike the conventional analog voltage or current mode circuits, our circuits use transient operation in charging/discharging processes to capacitors through resistors. Since the circuits calculate multiple weighted-sums by charging a capacitance, they can be operated with extremely low energy consumption. However, because a relatively long time constant is required to guarantee calculation resolution in the time domain, they have to use very high-resistance devices, on the order of giga-ohms. We designed, fabricated, and tested a proof-of-concept complementary metal-oxide-semiconductor (CMOS) VLSI chip using a 250-nm fabrication technology to verify weighted-sum operation based on the proposed model with binary weights and PWM input signals, which realizes the BinaryConnect model. In the chip, memory cells of static-random-access memory (SRAM) are used for synaptic connection weights. High-resistance operation was realized by using the subthreshold operation region of MOS transistors, unlike in the ordinary in-memory-computing circuits. We evaluated the energy efficiency and temperature characteristics by measurement using the fabricated chip, where the highest energy efficiency for the weighted-sum calculation was 300 TOPS/W (Tera-Operations Per Second per Watt). The effects by a temperature change can be compensated for by adjusting the bias voltage. If state-of-the-art VLSI technology is used to implement the proposed model, an energy efficiency of more than 1,000 TOPS/W will be possible.

12 February 2016 in Applied Physics Express
Verified
Journal Article
Open access
PDF file
TTTakashi Tohara
Takashi Tohara
HLHaichao Liang
Haichao Liang
HTHirofumi Tanaka
Hirofumi Tanaka
MIMakoto Igarashi
Makoto Igarashi
SSSeiji Samukawa
Seiji Samukawa
TMTakashi Morie
Takashi Morie
A nanodisk array connected with a fin field-effect transistor is fabricated and analyzed for spiking neural network applications. This nanodevice performs weighted sums in the time domain using rising slopes of responses triggered by input spike pulses. The nanodisk arrays, which act as a resistance of several giga-ohms, are fabricated using a self-assembly bio-nano-template technique. Weighted sums are achieved with an energy dissipation on the order of 1 fJ, where the number of inputs can be more than one hundred. This amount of energy is several orders of magnitude lower than that of conventional digital processors.

12 January 2024 in Communications Physics
Verified
Journal Article
Open access
PDF file
YSYusuke Sakemi
Yusuke Sakemi
SNSou Nobukawa
Sou Nobukawa
TMToshitaka Matsuki
Toshitaka Matsuki
TMTakashi Morie
Takashi Morie
KAKazuyuki Aihara
Kazuyuki Aihara
Reservoir computing (RC) can efficiently process time-series data by mapping the input signal into a high-dimensional space via randomly connected recurrent neural networks (RNNs), which are referred to as a reservoir. The high-dimensional representation of time-series data in the reservoir simplifies subsequent learning tasks. Although this simple architecture allows fast learning and facile physical implementation, the learning performance is inferior to that of other state-of-the-art RNN models. In this study, to improve the learning ability of RC, we propose self-modulated RC (SM-RC) that extends RC by adding a self-modulation mechanism. SM-RC can perform attention tasks where input information is retained or discarded depending on the input signal. We find that a chaotic state can emerge as a result of learning in SM-RC. Furthermore, we demonstrate that SM-RC outperforms RC in NARMA and Lorenz model tasks. Because the SM-RC architecture only requires two additional gates, it is physically implementable as RC, thereby providing a direction for realizing edge artificial intelligence.