Searcharxiv⌕ Search

arXiv subjects

Hiroshi Watanabe

Publications and source records attributed to Hiroshi Watanabe.

At least 55 records · Page 3Linked to original sources
arXiv2024

Fluctuation of a bilayer composed by amphipathic molecules

We investigated the properties of bilayer fluctuations using molecular dynamics. We modeled the amphipathic molecules as a diatomic molecule, constructed a bilayer in the solvent, and observed the Fourier spectrum of its fluctuations. The results showed that $q^4$ behavior was dominant at the high temperature, whereas a crossover from $q^4$ to $q^2$ was observed at the low temperature. This behavior is complementary to the results for monolayer membranes, where $q^2$ behavior was dominant when the interfacial tension was high, whereas a crossover from $q^2$ to $q^4$ was observed when the interfacial tension was lowered~[S. Kikuchi and H. Watanabe, J. Chem. Phys., $\bm{158}$ 12 (2023)]. These results suggest that the restoring force is dominated by elasticity at high temperatures and by interfacial tension at low temperatures. We also observed radial distribution functions to investigate the structure of amphipathic molecules in bilayers. The structure of the amphipathic molecules depends on temperature in the direction horizontal to the interface, while it does not depend on temperature in the direction perpendicular to the interface. This result suggests that bilayers increase fluctuations while maintaining the membrane structure in high temperature.

cond-mat.stat-mech↗
arXiv2024

Scalable Image Coding for Humans and Machines Using Feature Fusion Network

As image recognition models become more prevalent, scalable coding methods for machines and humans gain more importance. Applications of image recognition models include traffic monitoring and farm management. In these use cases, the scalable coding method proves effective because the tasks require occasional image checking by humans. Existing image compression methods for humans and machines meet these requirements to some extent. However, these compression methods are effective solely for specific image recognition models. We propose a learning-based scalable image coding method for humans and machines that is compatible with numerous image recognition models. We combine an image compression model for machines with a compression model, providing additional information to facilitate image decoding for humans. The features in these compression models are fused using a feature fusion network to achieve efficient image compression. Our method's additional information compression model is adjusted to reduce the number of parameters by enabling combinations of features of different sizes in the feature fusion network. Our approach confirms that the feature fusion network efficiently combines image compression models while reducing the number of parameters. Furthermore, we demonstrate the effectiveness of the proposed scalable coding method by evaluating the image compression performance in terms of decoded image quality and bitrate.

cs.CV↗
arXiv2024

Refining Coded Image in Human Vision Layer Using CNN-Based Post-Processing

Scalable image coding for both humans and machines is a technique that has gained a lot of attention recently. This technology enables the hierarchical decoding of images for human vision and image recognition models. It is a highly effective method when images need to serve both purposes. However, no research has yet incorporated the post-processing commonly used in popular image compression schemes into scalable image coding method for humans and machines. In this paper, we propose a method to enhance the quality of decoded images for humans by integrating post-processing into scalable coding scheme. Experimental results show that the post-processing improves compression performance. Furthermore, the effectiveness of the proposed method is validated through comparisons with traditional methods.

cs.CV↗
arXiv2024

Implicit Neural Representation for Videos Based on Residual Connection

Video compression technology is essential for transmitting and storing videos. Many video compression methods reduce information in videos by removing high-frequency components and utilizing similarities between frames. Alternatively, the implicit neural representations (INRs) for videos, which use networks to represent and compress videos through model compression. A conventional method improves the quality of reconstruction by using frame features. However, the detailed representation of the frames can be improved. To improve the quality of reconstructed frames, we propose a method that uses low-resolution frames as residual connection that is considered effective for image reconstruction. Experimental results show that our method outperforms the existing method, HNeRV, in PSNR for 46 of the 49 videos.

cs.CV↗
arXiv2024

Image Coding for Machines with Edge Information Learning Using Segment Anything

Image Coding for Machines (ICM) is an image compression technique for image recognition. This technique is essential due to the growing demand for image recognition AI. In this paper, we propose a method for ICM that focuses on encoding and decoding only the edge information of object parts in an image, which we call SA-ICM. This is an Learned Image Compression (LIC) model trained using edge information created by Segment Anything. Our method can be used for image recognition models with various tasks. SA-ICM is also robust to changes in input data, making it effective for a variety of use cases. Additionally, our method provides benefits from a privacy point of view, as it removes human facial information on the encoder's side, thus protecting one's privacy. Furthermore, this LIC model training method can be used to train Neural Representations for Videos (NeRV), which is a video compression model. By training NeRV using edge information created by Segment Anything, it is possible to create a NeRV that is effective for image recognition (SA-NeRV). Experimental results confirm the advantages of SA-ICM, presenting the best performance in image compression for image recognition. We also show that SA-NeRV is superior to ordinary NeRV in video compression for machines. Code is available at https://github.com/final-0/SA-ICM.

cs.CV↗
arXiv2024

Phase Repair for Time-Domain Convolutional Neural Networks in Music Super-Resolution

Audio Super-Resolution (SR) is an important topic as low-resolution recordings are ubiquitous in daily life. In this paper, we focus on the music SR task, which is challenging due to the wide frequency response and dynamic range of music. Many models are designed in time domain to jointly process magnitude and phase of audio signals. However, prior works show that approaches using Time-Domain Convolutional Neural Network (TD-CNN) tend to produce annoying artifacts in their waveform outputs, and the cause of the artifacts is yet to be identified. To the best of our knowledge, this work is the first to demonstrate the artifacts in TD-CNNs are caused by the phase distortion via a subjective experiment. We further propose Time-Domain Phase Repair (TD-PR), which uses a neural vocoder pre-trained on the wide-band data to repair the phase components in the waveform outputs of TD-CNNs. Although the vocoder and TD-CNNs are independently trained, the proposed TD-PR obtained better mean opinion score, significantly improving the perceptual quality of TD-CNN baselines. Since the proposed TD-PR only repairs the phase components of the waveforms, the improved perceptual quality in turn indicates that phase distortion has been the cause of the annoying artifacts of TD-CNNs. Moreover, a single pretrained vocoder can be directly applied to arbitrary TD-CNNs without additional adaptation. Therefore, we apply TD-PR to three TD-CNNs that have different architecture and parameter amount. Consistent improvements are observed when TD-PR is applied to all three TD-CNN baselines. Audio samples are available on the demo page.

cs.SD↗
arXiv2024

Unexpected linear conductivity in Landau-Zener model: limitations and improvements of the relaxation time approximation in the quantum master equation

The nonequilibrium steady states of quantum materials have many challenges. Here, we highlight issues with the relaxation time approximation (RTA) for the DC conductivity in insulating systems. The RTA to the quantum master equation (QME) is frequently employed as a simple method, yet this phenomenological approach is exposed as a fatal approximation, displaying metallic DC conductivity in insulating systems within the linear response regime. We find that the unexpected metallic behavior is caused by the fact that the density matrix in the RTA incompletely incorporates the first order of the external field. To solve this problem, we have derived a new calculation scheme based on the QME that ensure correct behavior in low electric fields. Our method reproduces well the overall features of the exact electric currents in the whole field region. It is not time-consuming, and its application to lattice systems is straightforward. This method will encourage progress in this research area as a simple way to more accurately describe nonequilibrium steady states.

cond-mat.mes-hall↗
arXiv2023

Energy Spectrum Analysis on a Red Blood Cell Model

It is important to understand the dynamics of red blood cells (RBCs) in blood flow. This requires the formulation of coarse-grained RBC models that reproduce the hydrodynamic properties of blood accurately. One of the models that successfully reproduce the rheology and morphology of blood has been proposed by Fedosov et al. [D. A. Fedosov, B. Caswell, and G. E. Karniadakis, Comput. Methods Appl. Mech. Eng., Vol. 199, 1937-1948 (2010)]. The proposed RBC model contains several parameters whose values are determined either by various experiments or physical requirements. In this study, we developed a new method of determining the parameter values precisely from the fluctuations of the RBC membrane. Specifically, we studied the relationship between the spectra of the fluctuations and model parameters. Characteristic peaks were observed in the spectra, whose peak frequencies were dependent on the parameter values. In addition, we investigated the spectra of the radius of gyration. We identified the peaks originating from the spring potential and the volume-conserving potential appearing in the spectra. These results lead to the precise experimental determination of the parameters used in the RBC model.

cond-mat.soft↗
arXiv2023

BigWavGAN: A Wave-To-Wave Generative Adversarial Network for Music Super-Resolution

Generally, Deep Neural Networks (DNNs) are expected to have high performance when their model size is large. However, large models failed to produce high-quality results commensurate with their scale in music Super-Resolution (SR). We attribute this to that DNNs cannot learn information commensurate with their size from standard mean square error losses. To unleash the potential of large DNN models in music SR, we propose BigWavGAN, which incorporates Demucs, a large-scale wave-to-wave model, with State-Of-The-Art (SOTA) discriminators and adversarial training strategies. Our discriminator consists of Multi-Scale Discriminator (MSD) and Multi-Resolution Discriminator (MRD). During inference, since only the generator is utilized, there are no additional parameters or computational resources required compared to the baseline model Demucs. Objective evaluation affirms the effectiveness of BigWavGAN in music SR. Subjective evaluations indicate that BigWavGAN can generate music with significantly high perceptual quality over the baseline model. Notably, BigWavGAN surpasses the SOTA music SR model in both simulated and real-world scenarios. Moreover, BigWavGAN represents its superior generalization ability to address out-of-distribution data. The conducted ablation study reveals the importance of our discriminators and training strategies. Samples are available on the demo page.

cs.SD↗
arXiv2023

Point Cloud Denoising and Outlier Detection with Local Geometric Structure by Dynamic Graph CNN

The digitalization of society is rapidly developing toward the realization of the digital twin and metaverse. In particular, point clouds are attracting attention as a media format for 3D space. Point cloud data is contaminated with noise and outliers due to measurement errors. Therefore, denoising and outlier detection are necessary for point cloud processing. Among them, PointCleanNet is an effective method for point cloud denoising and outlier detection. However, it does not consider the local geometric structure of the patch. We solve this problem by applying two types of graph convolutional layer designed based on the Dynamic Graph CNN. Experimental results show that the proposed methods outperform the conventional method in AUPR, which indicates outlier detection accuracy, and Chamfer Distance, which indicates denoising accuracy.

cs.CV↗
arXiv2023

Image Coding for Machines with Object Region Learning

Compression technology is essential for efficient image transmission and storage. With the rapid advances in deep learning, images are beginning to be used for image recognition as well as for human vision. For this reason, research has been conducted on image coding for image recognition, and this field is called Image Coding for Machines (ICM). There are two main approaches in ICM: the ROI-based approach and the task-loss-based approach. The former approach has the problem of requiring an ROI-map as input in addition to the input image. The latter approach has the problems of difficulty in learning the task-loss, and lack of robustness because the specific image recognition model is used to compute the loss function. To solve these problems, we propose an image compression model that learns object regions. Our model does not require additional information as input, such as an ROI-map, and does not use task-loss. Therefore, it is possible to compress images for various image recognition models. In the experiments, we demonstrate the versatility of the proposed method by using three different image recognition models and three different datasets. In addition, we verify the effectiveness of our model by comparing it with previous methods.

cs.CV↗
arXiv2023

VVC Extension Scheme for Object Detection Using Contrast Reduction

In recent years, video analysis using Artificial Intelligence (AI) has been widely used, due to the remarkable development of image recognition technology using deep learning. In 2019, the Moving Picture Experts Group (MPEG) has started standardization of Video Coding for Machines (VCM) as a video coding technology for image recognition. In the framework of VCM, both higher image recognition accuracy and video compression performance are required. In this paper, we propose an extention scheme of video coding for object detection using Versatile Video Coding (VVC). Unlike video for human vision, video used for object detection does not require a large image size or high contrast. Since downsampling of the image can reduce the amount of information to be transmitted. Due to the decrease in image contrast, entropy of the image becomes smaller. Therefore, in our proposed scheme, the original image is reduced in size and contrast, then coded with VVC encoder to achieve high compression performance. Then, the output image from the VVC decoder is restored to its original image size using the bicubic method. Experimental results show that the proposed video coding scheme achieves better coding performance than regular VVC in terms of object detection accuracy.

cs.CV↗
arXiv2023

Accuracy Improvement of Object Detection in VVC Coded Video Using YOLO-v7 Features

With advances in image recognition technology based on deep learning, automatic video analysis by Artificial Intelligence is becoming more widespread. As the amount of video used for image recognition increases, efficient compression methods for such video data are necessary. In general, when the image quality deteriorates due to image encoding, the image recognition accuracy also falls. Therefore, in this paper, we propose a neural-network-based approach to improve image recognition accuracy, especially the object detection accuracy by applying post-processing to the encoded video. Versatile Video Coding (VVC) will be used for the video compression method, since it is the latest video coding method with the best encoding performance. The neural network is trained using the features of YOLO-v7, the latest object detection model. By using VVC as the video coding method and YOLO-v7 as the detection model, high object detection accuracy is achieved even at low bit rates. Experimental results show that the combination of the proposed method and VVC achieves better coding performance than regular VVC in object detection accuracy.

cs.CV↗
arXiv2023

Observation of Electronic Structure Modification in the Hidden Order Phase of CeCoSi

CeCoSi with no local inversion symmetric crystal structure ($P4/nmm$) exhibits a phase transition of unknown origin (Hidden Order: HO) at about 12 K ($T_0$) above the antiferromagnetic transition temperature ($T_{\rm N}=9.4$ K). The electronic structure change across $T_0$ was investigated with high-precision optical reflection spectroscopy. The optical spectrum changed from a typical metallic behavior above $T_0$ to a gap-like structure at around 15 meV below $T_0$. The gap-like structure was unchanged across $T_{\rm N}$ except for the narrowing of the Drude component of carriers due to the suppression of magnetic fluctuations. This result suggests a slight change from the typical metallic electronic structure above $T_0$ to that with an energy gap near the Fermi level in the HO phase. The change in electronic structure in the HO phase was concluded to be due to electron/valence instability.

cond-mat.str-el↗
arXiv2023

Monte Carlo study of cuprate superconductors in a four-band $d$-$p$ model: Role of orbital degrees of freedom

Understanding the complex phase diagram of cuprate superconductors is a long-standing challenging problem. Recent studies have shown that orbital degrees of freedom, both Cu $e_g$ orbitals and O $p$ orbitals, are a key ingredient for a unified understanding of cuprate superconductors, including the material dependence. Here we investigate a four-band $d$-$p$ model derived from the first-principles calculations with the variational Monte Carlo method, which allows us to elucidate competing orders on an equal footing. The obtained results can consistently explain the doping dependence of superconductivity, antiferromagnetic and stripe phases, phase separation in the underdoped region, and also novel magnetism in the heavily-overdoped region. Our four-band $d$-$p$ model with neighbouring intersite interactions is a minimal model to describe the phase diagram comprehensively. The presence of $p$ orbitals is critical to the charge-stripe features, which induce two types of stripe phases with $s'$-wave and $d$-wave bond stripe. On the other hand, the presence of $d_{z^2}$ orbital is indispensable to material dependence of the superconducting transition temperature ($T_{\mathrm{c}}$), and enhances local magnetic moment as a source of novel magnetism in the heavily-overdoped region as well. These findings beyond one-band description could provide a major step toward a full explanation of unconventional normal state and high $T_{\mathrm{c}}$ in cuprate supercondutors.

cond-mat.supr-con↗
arXiv2023

Effect of surfactants on the elasticity of the liquid--liquid interface

We investigated the effect of surfactants on an interface between two kinds of liquid by molecular dynamics simulation. We adopted the simple bead-spring model with two atoms as the surfactants. We controlled the interfacial tension of the surfactant adsorbed on the interface by changing the bond length. Although the interface's structure changed depending on the magnitude of the interfacial tension, the interface was stable even under conditions where the interfacial tension was virtually zero. The Fourier spectrum of the fluctuations of the surface structure showed a crossover from $q^2$ to $q^4$ when the interfacial tension was almost zero, where $q$ is the wavenumber. This crossover means that the bending rigidity is dominant for the restoring force when the surfactant molecules are sufficiently absorbed on the interface and the interfacial tension is almost zero, whereas the interfacial tension is dominant when the interfacial tension is a finite value.

cond-mat.stat-mech↗
arXiv2023

Current-Induced Metallization and Valence Transition in Black SmS

A strongly-correlated insulator, samarium mono-sulfide (SmS), presents not only the pressure-induced insulator-to-metal transition (IMT) with the color change from black to golden-yellow but also current-induced IMT (CIMT) with negative resistance. To clarify the origin of the CIMT of SmS, the electronic structure change has been investigated by optical reflectivity and angle-integrated photoelectron spectra by applying an electric current. At lower temperatures than about 100 K, where the nonlinear $V$-$I$ curve has been observed, the carrier density rapidly increases, accompanied by decreasing relaxation time of carriers with increasing current. Then, the direct gap size increases, and the mean valence changes from Sm$^{2+}$-dominant SmS to the mixed-valent one with increasing current. These results suggest that the CIMT originates from increasing the Sm $4f$-$5d$ hybridization intensity induced by the applied current.

cond-mat.str-el↗
arXiv2023

Non-monotonic behavior of the Binder Parameter in the discrete spin systems

We study a non-monotonic behavior of the Binder parameter, which appears in the discrete spin systems. We show that the Binder parameters of the Potts model are non-monotonic for $q=3$ and $4$, while they are monotonic for the Ising case ($q=2$). Using the Fortuin-Kasteleyn graph representation, we find that the improved estimator of the Binder parameter consists of two terms with values only in high- and low-temperature regions. The non-monotonic behavior is found to originate from the low-temperature term. With the appropriately defined order parameter, we can reduce the influence of the low-temperature term, and as a result, the non-monotonic behavior can also be reduced. We propose new definitions of the order parameter, which reduces or eliminates the non-monotonic behavior of the Binder parameter in a system for which the improved estimator of the Binder parameter is unknown.

cond-mat.stat-mech↗