no code implementations • 5 Jan 2024 • Ikumi Okubo, Keisuke Sugiura, Hiroki Matsutani
To mitigate the computational complexity, recently, a hybrid approach has been proposed, which uses ResNet as a backbone architecture and replaces a part of its convolution layers with an MHSA (Multi-Head Self-Attention) mechanism.
no code implementations • 23 Dec 2023 • Kazuki Sunaga, Keisuke Sugiura, Hiroki Matsutani
A graph embedding is an emerging approach that can represent a graph structure with a fixed-length low-dimensional vector.
no code implementations • 19 Dec 2022 • Takeya Yamada, Hiroki Matsutani
In this case, both the neural network retraining and the proposed concept drift detection are done only by sequential computation to reduce computation cost and memory utilization.
no code implementations • 19 Aug 2022 • Yuto Hoshino, Hiroki Kawakami, Hiroki Matsutani
Furthermore, we show that our approach can reduce communication size by up to 92. 4% compared with a baseline ResNet model using CIFAR-10 dataset.
1 code implementation • 2 Mar 2022 • Kazuki Sunaga, Masaaki Kondo, Hiroki Matsutani
This article introduces a neural network based on-device learning (ODL) approach to address this issue by retraining in deployed environments.
no code implementations • 26 Oct 2021 • Masaki Furukawa, Hiroki Matsutani
In distributed reinforcement learning, Actor nodes acquire experiences by interacting with a given environment and a Learner node optimizes their DQN model.
no code implementations • 27 Jul 2021 • Hiroki Kawakami, Hirohisa Watanabe, Keisuke Sugiura, Hiroki Matsutani
It is implemented on Xilinx ZCU104 board and evaluated in terms of domain adaptation accuracy, inference speed, FPGA resource utilization, and speedup rate compared to a software counterpart.
no code implementations • 17 Mar 2021 • Mineto Tsukada, Hiroki Matsutani
Currently there has been increasing demand for real-time training on resource-limited IoT devices such as smart sensors, which realizes standalone online adaptation for streaming data without data transfers to remote servers.
no code implementations • 31 Dec 2020 • Hirohisa Watanabe, Hiroki Matsutani
In this paper, using Euler method as an ODE solver, a part of ODENet is implemented as a dedicated logic on a low-cost FPGA (Field-Programmable Gate Array) board, such as PYNQ-Z2 board.
no code implementations • 29 May 2020 • Keisuke Sugiura, Hiroki Matsutani
In this paper, we propose a resource-efficient FPGA implementation for accelerating scan matching computations, which typically cause a major bottleneck in 2D LiDAR SLAM methods.
no code implementations • 10 May 2020 • Hirohisa Watanabe, Mineto Tsukada, Hiroki Matsutani
In addition, we propose a combination of L2 regularization and spectral normalization for the on-device reinforcement learning so that output values of the neural network can be fit into a certain range and the reinforcement learning becomes stable.
no code implementations • 27 Feb 2020 • Rei Ito, Mineto Tsukada, Hiroki Matsutani
We extend it for an on-device federated learning so that edge devices can exchange their trained results and update their model by using those collected from the other edge devices.
no code implementations • 23 Jul 2019 • Mineto Tsukada, Masaaki Kondo, Hiroki Matsutani
However, (1) the iterative optimization often requires significant efforts to follow changes in the distribution of normal data (i. e., concept drift), and (2) data transfers between edge and server impose additional latency and energy consumption.
Semi-supervised Anomaly Detection Supervised Anomaly Detection +1