arXiv subjects

Masaki Saito

Publications and source records attributed to Masaki Saito.

8 recordsLinked to original sources

arXiv2025

Extended Diffeomorphism for Real-Time Motion Replication in Workspaces with Different Spatial Arrangements

This paper presents two types of extended diffeomorphism designs to compensate for spatial placement differences between robot workspaces. Teleoperation of multiple robots is attracting attention to expand the utilization of the robot embodiment. Real-time reproduction of robot motion would facilitate the efficient execution of similar tasks by multiple robots. A challenge in the motion reproduction is compensating for the spatial arrangement errors of target keypoints in robot workspaces. This paper proposes a methodology for smooth mappings that transform primary robot poses into follower robot poses based on the predefined key points in each workspace. Through a picking task experiment using a dual-arm UR5 robot, this study demonstrates that the proposed mapping generation method can balance lower mapping errors for precise operation and lower mapping gradients for smooth replicated movement.

cs.RO↗

arXiv2024

PolMERLIN: Self-Supervised Polarimetric Complex SAR Image Despeckling with Masked Networks

Despeckling is a crucial noise reduction task in improving the quality of synthetic aperture radar (SAR) images. Directly obtaining noise-free SAR images is a challenging task that has hindered the development of accurate despeckling algorithms. The advent of deep learning has facilitated the study of denoising models that learn from only noisy SAR images. However, existing methods deal solely with single-polarization images and cannot handle the multi-polarization images captured by modern satellites. In this work, we present an extension of the existing model for generating single-polarization SAR images to handle multi-polarization SAR images. Specifically, we propose a novel self-supervised despeckling approach called channel masking, which exploits the relationship between polarizations. Additionally, we utilize a spatial masking method that addresses pixel-to-pixel correlations to further enhance the performance of our approach. By effectively incorporating multiple polarization information, our method surpasses current state-of-the-art methods in quantitative evaluation in both synthetic and real-world scenarios.

cs.CV↗

arXiv2020

Electronic States of Al-Mg-Zn Quasicrystal and Its Approximant based on the First-Principles Calculations

First-principles calculations are performed to investigate the electronic states of 1/1 and 2/1 approximants with the composition Al$_{15}$Zn$_{40}$Mg$_{45}$ which is close to the quasicrystal (QC) Al$_{14.9}$Zn$_{41.0}$Mg$_{44.1}$ in which the superconductivity is recently discovered. The density of states for the 1/1 approximant shows a wide pseudogap structure near the Fermi level as commonly observed in various approximants of the QCs, whereas those for the 2/1 approximant do not show such a wide pseudogap. Instead of the wide pseudogap, the 2/1 approximant shows a remarkable narrow pseudogap at the Fermi level in contrast to the 1/1 approximant, which shows a shallow hump at the Fermi level within the wide pseudogap. This seems to be consistent with the experimental observation of the electrical resistivity, which increases with decreasing temperature from the room temperature down to around the superconducting transition temperature for the 2/1 approximant together with the QC, whereas it monotonically decreases for the 1/1 approximant.

cond-mat.supr-con↗

arXiv2020

Train Sparsely, Generate Densely: Memory-efficient Unsupervised Training of High-resolution Temporal GAN

Training of Generative Adversarial Network (GAN) on a video dataset is a challenge because of the sheer size of the dataset and the complexity of each observation. In general, the computational cost of training GAN scales exponentially with the resolution. In this study, we present a novel memory efficient method of unsupervised learning of high-resolution video dataset whose computational cost scales only linearly with the resolution. We achieve this by designing the generator model as a stack of small sub-generators and training the model in a specific way. We train each sub-generator with its own specific discriminator. At the time of the training, we introduce between each pair of consecutive sub-generators an auxiliary subsampling layer that reduces the frame-rate by a certain ratio. This procedure can allow each sub-generator to learn the distribution of the video at different levels of resolution. We also need only a few GPUs to train a highly complex generator that far outperforms the predecessor in terms of inception scores.

cs.CV↗

arXiv2019

1/f^2 spectra of decoherence noise on ^75^As nuclear spins in bulk GaAs

To identify the decoherence origin, frequency spectra using multiple π-pulses have been extensively studied. However, little has been discussed on how to define the spectral intensities from multiple-echo decays and how to incorporate the Hahn-echo T_2 in the noise spectra. Here, we show that experiments based on two theories solve these issues. With the previous theory clarifying that the spectral intensity should be given as the decay in the long-time limit, the intensity can be deduced without experimental artifacts usually entailed in the initial process. The other is the fluctuation-dissipation theory, with which the Hahn-echo T_2 is utilized as the zero-frequency limit of the noise spectrum and as an answer to the divergent issue on the 1/f^n noises. As a result, arsenic nuclear spins are found to exhibit 1/f2 dependences over two orders of magnitude in all the substrates of un-doped, Cr-doped semi-insulating and Si-doped metallic GaAs at 297 K. The 1/f^2 dependence indicates single noise source that is characterized by the characteristic frequency f_c_^un^=170 Hz, f_c_^Cr^=210 Hz and f_c_^Si^ =460 Hz. These f_c values are explained by a model that the decoherence is caused by the fluctuations of next-nearest-neighboring nuclear spins.

cond-mat.mes-hall↗

arXiv2017

ChainerCV: a Library for Deep Learning in Computer Vision

Despite significant progress of deep learning in the field of computer vision, there has not been a software library that covers these methods in a unifying manner. We introduce ChainerCV, a software library that is intended to fill this gap. ChainerCV supports numerous neural network models as well as software components needed to conduct research in computer vision. These implementations emphasize simplicity, flexibility and good software engineering practices. The library is designed to perform on par with the results reported in published papers and its tools can be used as a baseline for future research in computer vision. Our implementation includes sophisticated models like Faster R-CNN and SSD, and covers tasks such as object detection and semantic segmentation.

cs.CV↗

arXiv2017

Temporal Generative Adversarial Nets with Singular Value Clipping

In this paper, we propose a generative model, Temporal Generative Adversarial Nets (TGAN), which can learn a semantic representation of unlabeled videos, and is capable of generating videos. Unlike existing Generative Adversarial Nets (GAN)-based methods that generate videos with a single generator consisting of 3D deconvolutional layers, our model exploits two different types of generators: a temporal generator and an image generator. The temporal generator takes a single latent variable as input and outputs a set of latent variables, each of which corresponds to an image frame in a video. The image generator transforms a set of such latent variables into a video. To deal with instability in training of GAN with such advanced networks, we adopt a recently proposed model, Wasserstein GAN, and propose a novel method to train it stably in an end-to-end manner. The experimental results demonstrate the effectiveness of our methods.

cs.LG↗

arXiv2002

Experimental and Theoretical Study of the Residual Product Nuclide Yields in 100-2600 MeV Proton-Irradiated Thin Targets

The work is aimed at experimental determining and computer simulating the independent and cumulative yields of residual product nuclei in the target and structure materials of the transmutation facilities driven by high-current accelerators. The ITEP U-10 accelerator was used in 48 experiments to obtain more than 4000 values of the yields of radioactive residual product nuclei in 0.1-2.6 GeV proton-irradiated thin 182,183,184,186-W, nat-W, 56-Fe, 58-Ni, 93-Nb, 232-Th, nat-U, 99-Tc, 59-Co 63,65-Cu, nat-Hg, 208-Pb, and 27-Al targets. The results of verifying the LAHET, CEM95, CEM2k, CASCADE, CASCADE/INPE, YIELDX, HETC, INUCL, and other simulation codes are presented.

nucl-ex↗