記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

To use reservoir computing (RC) for practical tasks, both a high memory capacity and nonlinearity are required; however, some RC models have the problem of a low memory capacity. We propose a delay mechanism for increasing the memory capacity in RC as well as a simple and small-scale digital circuit for implementing the delay mechanism. The proposed delay mechanism is integrated into the input layer of the RC model and is expected to be implemented in several RC models, such as material reservoirs and chaotic Boltzmann machine (CBM)-RC. We conducted experiments using a CBM-RC with a delay mechanism (CBM-RC-DL) and evaluated the performance improvement achieved by introducing a delay mechanism. We used CBM-RC as the base model because it is an appropriate model for the hardware implementation of large networks but has a low memory capacity. The experimental results for CBM-RC-DL indicated that the delay mechanism significantly increased the memory capacity of CBM-RC with the addition of a small-scale circuit. Furthermore, the entire synthesized CBM-RC-DL was sufficiently small-scale to be implemented in a field-programmable gate array for edge computing, and it outperformed conventional methods in nonlinear autoregressive moving average 10 (NARMA10) - a benchmark task for time-series data processing. The proposed delay mechanism can facilitate the use of many RC models because of its simple structure.

DOI： 10.1109/ISCAS58744.2024.10557841

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85198500940&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

This study proposes a low computational cost hand-waving action recognition system for non-verbal communication in home service robots. The system is based on an echo state network, which requires lower computational costs than that of deep neural networks (DNNs), and processes time-series data of skeletal coordinates of humans to recognize hand-waving actions. Additionally, this study proposes and compares two types of Preprocessing ing methods of the skeletal coordinates to ensure the robustness of the human positions on the frame: one method extracts shoulder and arm angles, which are invariable regardless of the humans’ positions and the other normalizes the skeletal coordinates. The experimental result shows that the proposed system has competitive accuracy and is robust to varying human positions.

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85190300153&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

The task planning system is required to accomplish various requests from a human in real-world environments. SayCan, one of the task planning systems, has high accuracy. However, its accuracy decreases for requests that include abstract nouns of the ambiguous word/phrase. We propose a novel task planning system based on SayCan that introduces a function for checking concrete names of abstract nouns and a rule-based skill extraction, enhancing accuracy. The proposed system facilitates the interpretation of requests and enables appropriate task planning. The effectiveness of the proposed system was demonstrated at RoboCup@Home, where it achieved high performance.

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85190283813&origin=inward

記述言語：英語   掲載種別：研究論文（研究会，シンポジウム資料等）

記述言語：英語   掲載種別：研究論文（学術雑誌）

The artificial intelligence (AI) paradigm shifts from software to implementing general-purpose or application-specific hardware systems with lower power requirements. This study explored a material physical reservoir consisting of a material random network, called in-materio physical reservoir computing (RC) to achieve efficient hardware systems. The device, made up of a random, highly interconnected network of nonlinear Ag2Se nanojunctions as reservoir nodes, demonstrated the requisite characteristics of an in-materio physical reservoir, including but not limited to nonlinear switching, memory, and higher harmonic generation. The power consumption of the in-materio physical reservoir is 0.07 nW per nanojunctions, confirming its highly efficient information processing system. As a hardware reservoir, the devices successfully performed waveform generation tasks. Finally, a voice classification by an in-materio physical reservoir is achieved over 80%, comparable to an RC software simulation. In-materio physical RC with rich nonlinear dynamics has huge potential for next-generation hardware-based AI.

DOI： 10.1002/aelm.202400443

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85210391008&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

We propose look-up tables networks-based reservoir computing (LUTNet-RC). This work is the first trial of applying LUTNets to RC. LUTNet-RC consists of a LUT-based reservoir layer and a non-LUT-based output layer. LUTNets have disadvantages such as limited sparse connectivity and weights cannot be changed after implementation. However, when applied to a reservoir layer of RC (LUT-based reservoir layer), these disadvantages are eliminated, because this layer works with sparse connectivity and the weights are fixed, so only the advantage of small circuit resources is obtained. For the LUT-based reservoir layer, we propose and model a multi-bit weight reservoir, modifying the conventional binarized reservoir to improve calculation accuracy. In the case of LUTNets, the proposed multi-bit weight reservoir can be implemented without the increase in utilized circuit resources because LUTNets focus only on the input-output relationship on neurons. Additionally, we propose a speed-up method in the output layer with time division calculation, which compares the current network state with previous states and then calculates only status-changed neurons. As a result, we implement a LUTNet-RC with 1500 reservoir neurons on a field-programmable gate array (KR260) running at 100MHz. The utilized circuit resources are dominated by LUTs, which use approximately 26% of the total amount of LUTs. The LUTNet-RC can infer more than 106 data per second. We also verify the LUTNet-RC performance using nonlinear auto-regressive moving average 10 (NARMA10) and the performance is comparable to conventional works. We conclude that the LUTNet-RC is one of the highest-performance RC on an FPGA.

DOI： 10.1109/ICFPT59805.2023.00024

Kyutacar

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85187557709&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

We demonstrate a chaotic Boltzmann machine annealer, which is a field-programmable gate array (FPGA)-based high-performance annealing machine, and a control system powered by two open-source software packages. The annealer is employed in solving a large-scale real-world optimization problem, the traffic flow optimization. This involves distributing 500 car routes between the Haneda Airport area to the Pacifico Yokohama area, the venue of the International Conference on Field Programmable Technology 2023, while minimizing the total driving distance and preventing traffic jams. We obtain solutions that are comparable in accuracy to solutions of simulated annealing running on a graphics processing unit (GPU-SA) and a central processing unit (CPU-SA). The annealing machine on an FPGA is approximately 571 and 97600 times as fast as the GPUSA and the CPU-SA, respectively. We will demonstrate solving traffic flow optimization in the towns familiar to the visitors.

DOI： 10.1109/ICFPT59805.2023.00038

Kyutacar

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85187557051&origin=inward

担当区分：筆頭著者,　責任著者   記述言語：英語   掲載種別：研究論文（学術雑誌）

This study proposes a reservoir computing model with multiple readouts and an associated training method to enhance the training capability of reservoir computing. This study conducts a speaker classification task and a word classification task using an audio dataset consisting of digits pronounced by six persons. The experimental results reveal that the proposed model with multiple readouts outperforms a conventional model with a single readout.

DOI： 10.34385/proc.76.c3l-11

Kyutacar

CiNii Research

その他リンク： https://kyutech.repo.nii.ac.jp/records/2000241

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

Kyutacar

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

Ising machines are attracting attention for their ability to solve large-scale combinatorial optimization problems because these problems are difficult to solve. To accelerate the computing of Ising machines, implementation of Ising machines with digital circuits such as simulated annealing (SA) machines is in progress. However, these Ising machines on digital circuits require random number generators, which are implemented with large circuit resources. This work focuses on chaotic Boltzmann machines (CBMs), which imitate the stochastic behavior of Boltzmann machines (BMs) with deterministic chaotic dynamics. CBMs are one of the models that work as chaotic simulated annealing (CSA) machines within Ising machines. Therefore, we can implement the Ising machines without random number generators by using CBMs. In conventional work, CSA machines using CBMs (CBM-CSAs) are implemented with some hardware-oriented algorithms, but the CBM-CSA circuit is not optimized for these hardware-oriented algorithms. In the conventional CBM-CSA circuit, memory circuits are implemented separately, which prevents making the CBM-CSA from larger, and neuron circuits require the reset of accumulated values, which causes the increase in the calculation time. To solve these problems, we implement only one large memory circuit to make the CBM-CSA larger and improve the neuron circuits to allow dynamic changes of inputs to arithmetic circuits to inhibit the increase in the calculation time. As a result, we implement a CBM-CSA with 4096 nodes on an FPGA (Alveo U250), and the CBM-CSA can control 16-bit width weights and run at 100MHz. We evaluate the implemented CBM-CSA by solving K4000, max-cut problem, which is one of the combinatorial optimization problems. The best solution of CBM-CSA is comparable to that of the SA on the central processing unit (CPU). Moreover, the CBM-CSA is approximately 600 times as fast as the SA on the CPU and approximately twice as fast as the conventional Ising machine on an FPGA based on the improvements in this work. Furthermore, this work implements one of the highest-performance Ising machines on a single FPGA.

DOI： 10.1109/IJCNN54540.2023.10191342

Kyutacar

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85169581320&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

Emerging non-volatile memory devices, such as resistive random access memory (ReRAM) and voltage-controlled magnetoresistive random access memory (VC-MRAM), promise low energy consumption for artificial intelligence applications. However, when implementing deep neural networks (DNNs) using such memory devices, write-error may cause millions of bit-flipping to DNN. This easily degrades the DNN performance. To address this problem, we propose a novel repetition coding for deep-learning (RC-DL), which is a repetition coding designed to protect IEEE 32-bit floating-point (FP32) DNN models. Compared to conventional repetition coding, the proposed RC-DL exploits FP32 non-uniform magnitude encoding by increasing the repeat rates to protect sensitive bit positions and reduce the repeat rates to insensitive bit positions. Hence, RC-DL uses a number of bits equivalent to a 3-bit repetition code while delivering the performance close to 11-bit repetition code. We perform extensive Monte Carlo simulations to simulate the write-error property with ImageNet 2012 pretrained models. The DNN models with RC-DL are shown to be operable in the extremely imperfect environment while delivering with only minor reductions in DNN performance.

DOI： 10.1109/IJCNN54540.2023.10191433

Kyutacar

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85169554268&origin=inward

担当区分：筆頭著者,　責任著者   記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

This study aims to implement a reservoir-based convolutional neural network (CNN) on physical reservoir computing (RC) to develop an efficient image recognition system for edge AI. Therefore, we propose a novel reservoir-based convolution circuit system that uses in-material reservoir computing, a type of physical RC made from a sulfonated polyaniline network. The experimental results demonstrate that the proposed circuit system extracts image features in the same way as the original CNN and that a reservoir-based CNN on the in-material RC achieves an accuracy rate of 81.7% in an image classification task while an echo state network-based CNN achieves 87.7%.

DOI： 10.1109/ISCAS46773.2023.10181436

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85167720431&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

High-speed and accurate human recognition is necessary to realize safe autonomous mobile robots. Recently, human recognition methods based on deep learning have been studied extensively. However, these methods consume large amounts of power. Therefore, this study focuses on the Gaussian mixture model of multiresolution co-occurrence histograms of oriented gradients (GMM-MRCoHOG), which is a feature extraction method for human recognition that entails lower computational costs compared to deep learning-based methods, and aims to implement its hardware for high-speed, high-accuracy, and low-power human recognition. A digital hardware implementation method of GMM-MRCoHOG has been proposed. However, the method requires numerous look-up tables (LUTs) to store state spaces of GMM-MRCoHOG, thereby impeding the realization of human recognition systems. This study proposes a LUT reduction method to overcome this drawback by standardizing basis function arrangements of Gaussian mixture distributions in GMM-MRCoHOG. Experimental results show that the proposed method is as accurate as the previous method, and the memory required for state spaces consuming LUTs can be reduced to 1/504th of that required in the previous method.

DOI： 10.5220/0011698400003417

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85183596965&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

An Ag2Se nanowire random network was fabricated for application as a data augmentation device and combined with a binary convolutional neural network (BCNN) to achieve high accuracy in voice classification tasks. Due to the nonlinear high-dimensional characteristics resulting from the formation of the conductive filament at the cross junction, the Ag2Se device could transform input data into higher-order multiple signals, thereby enhancing the accuracy of the classification task by augmenting input signals. The results indicate that materials can realize data augmentation with the same performance as software, suggesting that material-based hardware can be used as an elemental technology for information processing.

DOI： 10.35848/1882-0786/acae6a

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85146494177&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

日本   北九州   2022年12月13日  -  2022年12月14日

担当区分：筆頭著者,　責任著者   記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

日本   北九州   2022年12月13日  -  2022年12月14日

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

日本   北九州   2022年12月13日  -  2022年12月14日

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

日本   北九州   2022年12月13日  -  2022年12月14日

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

日本   北九州   2022年12月13日  -  2022年12月14日

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

日本   北九州   2022年12月13日  -  2022年12月14日

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

担当区分：筆頭著者,　責任著者   記述言語：英語   掲載種別：研究論文（学術雑誌）

<p>Reservoir computing (RC) has attracted attention and has been used in many applications because of its low training cost. Multiple studies using RC for image recognition have been proposed, and some have achieved accuracy rates of greater than 99% on the MNIST dataset. For the Fashion-MNIST and CIFAR-10 datasets, however, they have not yet achieved high accuracy. This study proposes a novel convolutional neural network based on RC that can be optimized by ridge regression rather than back-propagation. The reservoir-based network has multiple reservoirs with various leak rates to extract features with various spatial frequencies from the inputs. The experimental results show that the performance of the proposed model achieves higher accuracy rates in the mentioned datasets compared with those of other reservoir-based image recognition approaches.</p>

DOI： 10.1587/nolta.13.397

Kyutacar

CiNii Research

その他リンク： http://hdl.handle.net/10228/00009117

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

In this research, we focus on Gaussian mixture model-multiresolution co-occurrence histograms of oriented gradients (GMM-MRCoHOG) features using luminance gradients in images and propose a hardware-oriented algorithm of GMM-MRCoHOG to implement it on a field programmable gate array (FPGA). The proposed method simplifies the calculation of luminance gradients, which is a high-cost operation in the conventional algorithm, by using lookup tables to reduce the circuit size. We also designed a human-detection digital architecture of the proposed algorithm for FPGA implementation using high-level synthesis. The verification results showed that the processing speed of the proposed architecture was approximately 123 times faster than that of the FPGA implementation of VGG-16.

DOI： 10.5220/0010848100003124

Kyutacar

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85182929508&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

Kyutacar

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

With the advent of deep learning, significant improvements in image recognition performance have been achieved. In image recognition, it is generally assumed that all the test data are composed of known classes. This approach is termed as closed-set recognition. In closed-set recognition, when an untrained, unknown class is input, it is recognized as one of the trained classes. The method whereby an unknown image is recognized as unknown when it is input is termed as open-set recognition. Although several open-set recognition methods have been proposed, none of these previous methods excel in terms of all three evaluation items: learning cost, recognition performance, and scalability from closed-set recognition models. To address this, we propose an open-set recognition method using the distance between features in the multidimensional feature space of neural networks. By applying center loss to the feature space, we aim to maintain the classification accuracy of closed-set recognition and improve the unknown detection performance. In our experiments, we achieved state-of-the-art performance on the MNIST, SVHN, and CIFAR-10 datasets. In addition, the proposed approach shows excellent performance in terms of the three evaluation items.

DOI： 10.5220/0010816600003124

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85143999490&origin=inward

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

Image classification assumes that all classes used in testing are known. Therefore, when an unknown class data is input, it cannot be recognized correctly. A method that enables unknown classes to be identified is called open set recognition. In this paper, we propose a method of open set recognition focusing on the feature space of the classifier and a Mahalanobis-based threshold. The experimental results show that the proposed method surpasses state-of-the-art methods on some datasets, demonstrating the potential of a method focusing on the feature space.

DOI： 10.1109/ISPACS51563.2021.9650985

Kyutacar

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85124169792&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

We propose a brain-inspired neural network model consisting of the hippocampus, prefrontal cortex, and amygdala models for a navigation system that acquires specific knowledge in home environments from few experiences. The proposed model was evaluated in a home environment using a robot simulator. In the experiment, the robot determines a path for navigation based on the knowledge acquired by the brain-inspired model.

DOI： 10.1109/ISPACS51563.2021.9651058

Kyutacar

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85124147021&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

A sulfonated polyaniline (SPAN) organic electrochemical network device (OEND) is fabricated using a simple drop-casting method on multiple Au electrodes for use in reservoir computing (RC). The SPAN network has humidity-dependent electrical properties. Under high humidity, the SPAN OEND exhibits mainly ionic conduction, including charging of an electric double layer and ionic diffusion. The nonlinearity and hysteresis of the current–voltage characteristics progressively increase with increasing humidity. The rich dynamic output behavior indicates wide variations for each electrode, which improves the RC performance because of the disordered network. For RC, waveform generation and short-term memory tasks are realized by a linear combination of outputs. The waveform task accuracy and memory capacity calculated from a short-term memory task reach 90% and 33.9, respectively. Improved spoken-digit classification is realized with 60% accuracy by only 12 outputs, demonstrating that the SPAN OEND can manage time series dynamic data operation in RC owing to a combination of rich dynamic and nonlinear electronic properties. The results suggest that SPAN-based electrochemical systems can be applied for material-based computing, by exploiting their intrinsic physicochemical behavior.

DOI： 10.1002/adma.202102688

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85114995164&origin=inward

担当区分：筆頭著者,　責任著者   記述言語：英語   掲載種別：研究論文（学術雑誌）

We propose a brain-inspired artificial intelligence model that acquires local knowledge from a small number of human-robot interactions, and uses this knowledge for future prediction. We focus on areas of the brain: the hippocampus, the amygdala, and the prefrontal cortex, thereby providing an internal representation of place and time, a value judgment, and a prediction. By integrating these functions, our model achieves local knowledge acquisition and future prediction. The proposed model is evaluated in a task for home service robots. In the experiment, the model learns local time-series events from a small number of interactions, and predicts the future events.

DOI： 10.34385/proc.74.a4l-c-4

CiNii Research

記述言語：英語   掲載種別：研究論文（学術雑誌）

We propose a robot navigation system using reinforcement learning and reservoir computing. Reservoir computing is a framework for the processing of complex time series based on a recurrent neural network (RNN). The proposed system is trained by reinforcement learning, which is a framework that performs trial and error based on the rewards given by the environment. The trained network generates motor commands reflecting on action values obtained by the reinforcement learning and operates the autonomous mobile robot.

DOI： 10.34385/proc.74.b3l-b-5

CiNii Research

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（学術雑誌）

This study develops an intelligent system for home service robots mimicking human brain function that can manage common knowledge applicable to any environment and local knowledge reflecting its specific environment. Deep learning is effective for acquiring common knowledge because the performance of deep learning relies on the amounts of training and big training data that can be accessed for such knowledge; however, deep learning is ineffective for acquiring local knowledge because no big training data for such knowledge exist. Thus, we propose a brain-inspired learning model and system for acquiring local knowledge using small training data. We focus on the amygdala because its classical fear conditioning is effective for training using small training data. We propose an amygdala-inspired classical conditioning model comprising multiple self-organizing maps (lateral nucleus) and a fully connected neural network (central nucleus), imitating the function and structure of the amygdala. The proposed model is applied to a task of a waiter robot in a restaurant, and the model can learn customers' preferences after only a few human-robot interactions. We accelerate the computation of the model and reduce its power consumption by proposing a hardware-oriented algorithm for the model and its digital hardware design and implement it in an XCZU9EG field programmable gate array. The hardware-oriented algorithm reduces the multiplication operations and exponential functions requiring huge hardware resources. The performance of the hardware operated at 150 MHz is 1,273 times faster than the software implementation on Arm Cortex-A53, and the power consumption of the chip is 5.009 W.

DOI： 10.1109/ACCESS.2020.3038161

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85097712435&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

記述言語：日本語   掲載種別：研究論文（研究会，シンポジウム資料等）

<p>Home service robots have begun attracting attention due to decreasing birthrate and increasing aging population. We,"Hibikino-Musashi@Home"team, develop home service robots, and participate international competitions; RoboCup@Home and World Robot Challenge (WRC), which are established to realize that robots will work and live with people in domestic environments. In this paper, we introduce our robots and technologies, especially an object recognition and manipulation system. We use You Only Look Once (YOLO) for object recognition. To train YOLO, a big dataset which is difficult to prepare in a short term is required. For this, we develop an automatic annotation system for YOLO which can generate 1.7 million training data, and we can complete training of YOLO in a day. To grasp objects, estimation of object orientations is required. Our system can detect the orientations by processing point clouds of objects. By using the system, we obtained about 30% of scores in Tidy Up Task in the WRC.</p>

DOI： 10.1299/jsmermd.2019.1P2-I02

CiNii Article

CiNii Research

その他リンク： https://ci.nii.ac.jp/naid/130007774525

担当区分：筆頭著者,　責任著者   記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

Deep neural networks (DNNs) have achieved state-of-the-art results in several computing tasks. However, the performance of these DNNs is reliant on the availability of large amounts of training data, which is not always present. We approached this problem by developing a brain-inspired amygdala model to achieve computer learning based on limited training data. The amygdala is an area of the brain associated with classical fear conditioning. The proposed amygdala model is composed of a single layer of deep self-organizing map network (deep SOM network) and a fully-connected neural network (FCNN), which imitates the function and structure of an amygdala. We applied the proposed amygdala model to a robot waiter task in a restaurant. The experimental results show that the model learned a customer's preferences after only a few human robot interactions. To develop the digital hardware of the amygdala model, we designed hardware for the deep SOM network and the FCNN and implemented them in an XCZU9EG field programmable gate array (FPGA). Our FPGA implementation of a deep SOM network with 272 neurons and an FCNN with three output neurons outperformed a software implementation on an Intel Core i5-3470 CPU by over 600 times.

DOI： 10.1109/ISCAS.2019.8702430

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85066804926&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

This live demonstration presents a VLSI chip based on “Time-domain Analog Computing with Transient states (TACT)” approach for intelligent processing on robots. This TACT chip, fabricated using 250-nm CMOS technology, implements a time-domain analog weighted-sum calculation model with very high energy efficiency. We integrate the TACT chip into a robot via Robot Operating System (ROS) interfaces. A human tracking robot demonstration is performed by the TACT chip with energy efficiency of 300 TOPS/W.

DOI： 10.1109/ISCAS.2019.8702222

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85066803614&origin=inward

担当区分：筆頭著者,　責任著者   記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

This live demonstration presents a brain-inspired amygdala model. Amygdala is an area of the brain that is associated with fear conditioning, which is a type of classical conditioning. The model can learn preferences through human-robot interactions by application of classical conditioning to the model. Additionally, to develop a high speed and low power system, we design a hardware of the amygdala model, and implemented the hardware into field programmable gate array.

DOI： 10.1109/ISCAS.2019.8702213

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85066814574&origin=inward

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

記述言語：日本語   掲載種別：研究論文（研究会，シンポジウム資料等）

<p>In real world applications of home service robots, they are required to be able to perform housekeeping or other tasks. "RoboCup@Home" is a competition for benchmarking home service robots. It includes tests to service robots for real world implementation. These tests require the combination of intelligent functions such as object detection and voice recognition. We have developed a home service robot "Exi@" and evaluated its performance by participating in RoboCup@Home. This paper describes real world implementation of robots and RoboCup@Home. The functions embedded in Exi@ and a task "Restaurant" are also explained.</p>

DOI： 10.1299/jsmermd.2018.2P2-E02

CiNii Article

CiNii Research

その他リンク： https://ci.nii.ac.jp/naid/130007552042

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

We aim to develop a recognition system of high accuracy and low power consumption by designing a digital circuit for deep neural networks (DNNs), and by implementing the circuit on field programmable gate arrays (FPGAs). DNNs include numerous multiply operations, whereas FPGAs include a limited number of multipliers. We aim to reduce the number of multiply operations generated by the algorithms within DNNs. Deep self-organizing map networks (DSNs) [1] are DNNs comprising self-organizing maps (SOMs) [2] as shown in Fig. 1. A hardware-oriented algorithm for SOMs has been proposed herein [3]. The algorithm represents SOMs by replacing multiply operations with bitshift operations. DSNs that include only a few multiply operations can then be represented by employing the algorithm. In this paper, we propose a hardware-oriented algorithm and a hardware architecture for DSNs. The hardware-oriented algorithm reduces multiply operations and exponential functions in a computation of SOM Module as shown in Fig. 1. In addition, we confirm that the algorithm does not worsen performance of DSN by a software simulation. Figure 2 shows error rates of DSN during learning MNIST Dataset [4]. The performance of the proposed algorithm achieve comparable results to the conventional algorithm. We also describe a DSN comprising three layers by Verilog-HDL as shown in Fig. 1, and implement it on a Xilinx Vertex-6 XC6VLX240T FPGA. Experimental results showed that the proposed DSN circuit estimates a label of an input image in 2 µs while the software implemented using an Intel Core i5-3470 (3.20 GHz) CPU estimates it in about 1 ms. Thus the hardware is 500 times faster than the software. Its logic utilization is shown in Table 1.

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85034254406&origin=inward

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

開催期間： 2024年03月01日 - 2024年03月03日   記述言語：英語  

開催期間： 2024年03月01日 - 2024年03月03日   記述言語：英語  

開催期間： 2024年03月01日 - 2024年03月03日   記述言語：英語  

開催期間： 2024年02月27日 - 2024年02月28日   記述言語：英語  

開催期間： 2023年12月09日   記述言語：日本語  

開催期間： 2023年11月20日 - 2023年11月21日   記述言語：英語  

開催期間： 2023年10月12日 - 2023年10月13日   記述言語：日本語  

開催期間： 2023年09月07日 - 2023年09月09日   記述言語：日本語  

開催期間： 2023年09月07日 - 2023年09月09日   記述言語：日本語  

開催期間： 2022年12月03日   記述言語：日本語  

開催期間： 2022年11月23日   記述言語：日本語  

開催期間： 2022年10月04日 - 2022年10月31日   記述言語：日本語  

開催期間： 2022年09月22日   記述言語：日本語  

開催期間： 2022年09月20日 - 2022年09月23日   記述言語：日本語  

開催期間： 2022年09月20日 - 2022年09月23日   記述言語：日本語   開催地：東北大学川内北キャンパス   国名：日本国  

開催期間： 2022年09月05日 - 2022年09月09日   記述言語：日本語  

開催期間： 2022年06月09日 - 2022年06月10日   記述言語：日本語  

開催期間： 2022年03月22日 - 2022年03月26日   記述言語：日本語  

開催期間： 2022年03月22日 - 2022年03月26日   記述言語：日本語  

開催期間： 2022年03月18日 - 2022年03月19日   記述言語：英語  

開催期間： 2022年03月18日 - 2022年03月19日   記述言語：英語  

開催期間： 2022年02月18日 - 2022年02月19日   記述言語：英語  

開催期間： 2021年12月05日 - 2021年12月08日   記述言語：英語  

開催期間： 2021年12月05日 - 2021年12月08日   記述言語：英語  

開催期間： 2021年12月03日   記述言語：日本語  

開催期間： 2021年09月10日 - 2021年09月13日   記述言語：日本語  

開催期間： 2021年09月08日 - 2021年09月11日   記述言語：日本語   開催地：オンライン   国名：日本国  

開催期間： 2021年03月16日 - 2021年03月19日   記述言語：日本語  

開催期間： 2020年12月12日 - 2020年12月19日   記述言語：英語  

開催期間： 2020年12月12日 - 2020年12月19日   記述言語：英語  

開催期間： 2020年10月   記述言語：日本語  

開催期間： 2019年09月04日 - 2019年09月06日   記述言語：日本語  

開催期間： 2018年09月   記述言語：日本語  

開催期間： 2018年06月   記述言語：日本語  

開催期間： 2017年09月   記述言語：日本語  

開催期間： 2017年09月   記述言語：英語  

開催期間： 2017年03月   記述言語：日本語  

開催期間： 2016年09月   記述言語：日本語  

開催期間： 2016年03月   記述言語：日本語  

開催期間： 2015年06月   記述言語：日本語  

出願番号：2021-007422  出願日：2021年01月20日

登録番号：特許第7551113号  登録日：2024年09月06日

出願番号：2021-007423  出願日：2021年01月20日

登録番号：特許第7551114号  登録日：2024年09月06日

発表言語：日本語   講演種別：その他   開催地：オンライン  

発表言語：日本語   講演種別：その他   開催地：オンライン  

発表言語：日本語   講演種別：その他  

発表言語：日本語   講演種別：招待講演  

発表言語：日本語   講演種別：座談会  

発表言語：英語   講演種別：招待講演  

受賞国：日本国

受賞国：フランス共和国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：タイ王国

受賞国：タイ王国

受賞国：タイ王国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：台湾

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：オーストラリア連邦

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：タイ王国

受賞国：日本国

受賞国：日本国

受賞国：カナダ

受賞国：カナダ

受賞国：カナダ

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：日本国

受賞国：日本国