Hiroki Kuroda
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Hiroki Kuroda has a total of 12 co-authors. Here are the top co-authors by number of shared publications:
- Daichi Kitahara
Ritsumeikan University
4 shared publications - Akira Hirabayashi
Dept. of Information Science and Engineering, Ritsumeikan University
2 shared publications Daichi Kitahara
Dept. of Information Science and Engineering, Ritsumeikan University
2 shared publications- Eiichi Yoshikawa
Colorado State University, Fort Collins, CO, USA
2 shared publications - Hiroshi Kikuchi
Center for Space Science and Radio Engineering, The University of Electro-Communications, Tokyo, Japan
2 shared publications - Masao Yamagishi
Department of Information and Communications Engineering, Tokyo Institute of Technology
2 shared publications
Latest Publications
- Open Access
- 18 August 2025
- Published by Elsevier in Journal of the Franklin Institute
- Vol. 362 (15)
A convex-nonconvex framework for enhancing minimization induced penalties
- 18 August 2025
- Published by Elsevier in Journal of the Franklin Institute
- Vol. 362 (15)
Open AccessA convex-nonconvex framework for enhancing minimization induced penalties
- Open Access
- 8 July 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in TechRxiv
Sharpening Minimization Induced Penalties
In order to remedy the underestimation of an advanced convex penalty defined via minimization, this paper proposes its nonconvex enhancement while preserving convexity of the associated regularized least-squares model. We first design a generalized Moreau enhanced minimization induced (GME-MI) penalty function by subtracting from the MI penalty its generalized Moreau envelope. Then, we derive an overall convexity condition for the GME-MI regularized least-squares model. Finally, under the overall convexity condition, characterizing the solution set of the GME-MI model with a carefully designed averaged nonexpansive operator, we develop a proximal splitting algorithm which is guaranteed to converge to a globally optimal solution. Numerical examples demonstrate the effectiveness of the proposed approach.
- 8 July 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in TechRxiv
Open AccessSharpening Minimization Induced Penalties
In order to remedy the underestimation of an advanced convex penalty defined via minimization, this paper proposes its nonconvex enhancement while preserving convexity of the associated regularized least-squares model. We first design a generalized Moreau enhanced minimization induced (GME-MI) penalty function by subtracting from the MI penalty its generalized Moreau envelope. Then, we derive an overall convexity condition for the GME-MI regularized least-squares model. Finally, under the overall convexity condition, characterizing the solution set of the GME-MI model with a carefully designed averaged nonexpansive operator, we develop a proximal splitting algorithm which is guaranteed to converge to a globally optimal solution. Numerical examples demonstrate the effectiveness of the proposed approach.
- Open Access
- 16 November 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 11, 128859-128874
Convex Estimation of Sparse-Smooth Power Spectral Densities From Mixtures of Realizations With Application to Weather Radar
In this paper, we propose a convex optimization-based estimation of sparse and smooth power spectral densities (PSDs) of complex-valued random processes from mixtures of realizations. While the PSDs are related to the magnitude of the frequency components of the realizations, it has been a major challenge to exploit the smoothness of the PSDs because penalizing the difference of the magnitude of the frequency components results in a nonconvex optimization problem that is difficult to solve. To address this challenge, we design the proposed model that jointly estimates the complex-valued frequency components and the nonnegative PSDs, which are respectively regularized to be sparse and sparse-smooth. By penalizing the difference of the nonnegative variable that estimates the PSDs, the proposed model can enhance the smoothness of the PSDs via convex optimization. Numerical experiments on the phased array weather radar, an advanced weather radar system, demonstrate that the proposed model achieves superior estimation accuracy compared to existing sparse estimation models, regardless of whether they are combined with a smoothing technique as a post-processing step or not.
- 16 November 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 11, 128859-128874
Open AccessConvex Estimation of Sparse-Smooth Power Spectral Densities From Mixtures of Realizations With Application to Weather Radar
In this paper, we propose a convex optimization-based estimation of sparse and smooth power spectral densities (PSDs) of complex-valued random processes from mixtures of realizations. While the PSDs are related to the magnitude of the frequency components of the realizations, it has been a major challenge to exploit the smoothness of the PSDs because penalizing the difference of the magnitude of the frequency components results in a nonconvex optimization problem that is difficult to solve. To address this challenge, we design the proposed model that jointly estimates the complex-valued frequency components and the nonnegative PSDs, which are respectively regularized to be sparse and sparse-smooth. By penalizing the difference of the nonnegative variable that estimates the PSDs, the proposed model can enhance the smoothness of the PSDs via convex optimization. Numerical experiments on the phased array weather radar, an advanced weather radar system, demonstrate that the proposed model achieves superior estimation accuracy compared to existing sparse estimation models, regardless of whether they are combined with a smoothing technique as a post-processing step or not.
- 1 April 2022
- Published by Institute of Electronics, Information and Communications Engineers (IEICE) in IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences
- Vol. E105.A (4) , 704-718
Image Super-Resolution via Generative Adversarial Networks Using Metric Projections onto Consistent Sets for Low-Resolution Inputs
This paper addresses single image super-resolution (SR) based on convolutional neural networks (CNNs). It is known that recovery of high-frequency components in output SR images of CNNs learned by the least square errors or least absolute errors is insufficient. To generate realistic high-frequency components, SR methods using generative adversarial networks (GANs), composed of one generator and one discriminator, are developed. However, when the generator tries to induce the discriminator's misjudgment, not only realistic high-frequency components but also some artifacts are generated, and objective indices such as PSNR decrease. To reduce the artifacts in the GAN-based SR methods, we consider the set of all SR images whose square errors between downscaling results and the input image are within a certain range, and propose to apply the metric projection onto this consistent set in the output layers of the generators. The proposed technique guarantees the consistency between output SR images and input images, and the generators with the proposed projection can generate high-frequency components with few artifacts while keeping low-frequency ones as appropriate for the known noise level. Numerical experiments show that the proposed technique reduces artifacts included in the original SR images of a GAN-based SR method while generating realistic high-frequency components with better PSNR values in both noise-free and noisy situations. Since the proposed technique can be integrated into various generators if the downscaling process is known, we can give the consistency to existing methods with the input images without degrading other SR performance.
- 1 April 2022
- Published by Institute of Electronics, Information and Communications Engineers (IEICE) in IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences
- Vol. E105.A (4) , 704-718
Image Super-Resolution via Generative Adversarial Networks Using Metric Projections onto Consistent Sets for Low-Resolution Inputs
This paper addresses single image super-resolution (SR) based on convolutional neural networks (CNNs). It is known that recovery of high-frequency components in output SR images of CNNs learned by the least square errors or least absolute errors is insufficient. To generate realistic high-frequency components, SR methods using generative adversarial networks (GANs), composed of one generator and one discriminator, are developed. However, when the generator tries to induce the discriminator's misjudgment, not only realistic high-frequency components but also some artifacts are generated, and objective indices such as PSNR decrease. To reduce the artifacts in the GAN-based SR methods, we consider the set of all SR images whose square errors between downscaling results and the input image are within a certain range, and propose to apply the metric projection onto this consistent set in the output layers of the generators. The proposed technique guarantees the consistency between output SR images and input images, and the generators with the proposed projection can generate high-frequency components with few artifacts while keeping low-frequency ones as appropriate for the known noise level. Numerical experiments show that the proposed technique reduces artifacts included in the original SR images of a GAN-based SR method while generating realistic high-frequency components with better PSNR values in both noise-free and noisy situations. Since the proposed technique can be integrated into various generators if the downscaling process is known, we can give the consistency to existing methods with the input images without degrading other SR performance.
- 7 March 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Signal Processing
- Vol. 70 (1053587X) , 1506-1520
Block-Sparse Recovery With Optimal Block Partition
This paper presents a convex recovery method for block-sparse signals whose block partitions are unknown a priori. We first introduce a nonconvex penalty function, where the block partition is adapted for the signal of interest by minimizing the mixed $\ell _{2}/\ell _{1}$ norm over all possible block partitions. Then, by exploiting a variational representation of the $\ell _{2}$ norm, we derive the proposed penalty function as a suitable convex relaxation of the nonconvex one. For a block-sparse recovery model designed with the proposed penalty, we develop an iterative algorithm which is guaranteed to converge to a globally optimal solution. Numerical experiments demonstrate the effectiveness of the proposed method.
- 7 March 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Signal Processing
- Vol. 70 (1053587X) , 1506-1520
Block-Sparse Recovery With Optimal Block Partition
This paper presents a convex recovery method for block-sparse signals whose block partitions are unknown a priori. We first introduce a nonconvex penalty function, where the block partition is adapted for the signal of interest by minimizing the mixed $\ell _{2}/\ell _{1}$ norm over all possible block partitions. Then, by exploiting a variational representation of the $\ell _{2}$ norm, we derive the proposed penalty function as a suitable convex relaxation of the nonconvex one. For a block-sparse recovery model designed with the proposed penalty, we develop an iterative algorithm which is guaranteed to converge to a globally optimal solution. Numerical experiments demonstrate the effectiveness of the proposed method.
- Open Access
- 3 March 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in TechRxiv
Block-Sparse Recovery with Optimal Block Partition
This paper presents a convex recovery method for block-sparse signals whose block partitions are unknown a priori. We first introduce a nonconvex penalty function, where the block partition is adapted for the signal of interest by minimizing the mixed l2/l1 norm over all possible block partitions. Then, by exploiting a variational representation of the l2 norm, we derive the proposed penalty function as a suitable convex relaxation of the nonconvex one. For a block-sparse recovery model designed with the proposed penalty, we develop an iterative algorithm which is guaranteed to converge to a globally optimal solution. Numerical experiments demonstrate the effectiveness of the proposed method.
- 3 March 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in TechRxiv
Open AccessBlock-Sparse Recovery with Optimal Block Partition
This paper presents a convex recovery method for block-sparse signals whose block partitions are unknown a priori. We first introduce a nonconvex penalty function, where the block partition is adapted for the signal of interest by minimizing the mixed l2/l1 norm over all possible block partitions. Then, by exploiting a variational representation of the l2 norm, we derive the proposed penalty function as a suitable convex relaxation of the nonconvex one. For a block-sparse recovery model designed with the proposed penalty, we develop an iterative algorithm which is guaranteed to converge to a globally optimal solution. Numerical experiments demonstrate the effectiveness of the proposed method.
- 24 February 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Geoscience and Remote Sensing
- Vol. 60 (01962892) , 1-19
Nonlinear Beamforming Based on Group-Sparsities of Periodograms for Phased Array Weather Radar
We propose nonlinear beamforming for phased array weather radars (PAWRs). Conventional beamforming is linear in the sense that a backscattered signal arriving from each elevation is reconstructed by a weighted sum of received signals, which can be seen as a linear transform for the received signals. For distributed targets such as raindrops, however, the number of scatterers is significantly large, differently from the case of point targets that are standard targets in array signal processing. Thus, the spatial resolution of the conventional linear beamforming is limited. To improve the spatial resolution, we exploit two characteristics of a periodogram of each backscattered signal from the distributed targets. The periodogram is a series of the powers of the discrete Fourier transform (DFT) coefficients of each backscattered signal and utilized as a nonparametric estimate of the power spectral density . Since each power spectral density is proportional to the Doppler frequency distribution, 1) major components of the periodogram are concentrated in the vicinity of the mean Doppler frequency and 2) frequency indices of the major components are similar between adjacent elevations. These are expressed as group-sparsities of the DFT coefficient matrix of the backscattered signals, and we propose to reconstruct the signals through convex optimization exploiting the group-sparsities. We consider two optimization problems. One problem roughly evaluates the group-sparsities and is relatively easy to solve. The other evaluates the group-sparsities more accurately but requires more time to solve. Both problems are solved with the alternating direction method of multipliers (ADMM) including nonlinear mappings. Simulations using synthetic and real-world PAWR data show that the proposed method dramatically improves the spatial resolution.
- 24 February 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Geoscience and Remote Sensing
- Vol. 60 (01962892) , 1-19
Nonlinear Beamforming Based on Group-Sparsities of Periodograms for Phased Array Weather Radar
We propose nonlinear beamforming for phased array weather radars (PAWRs). Conventional beamforming is linear in the sense that a backscattered signal arriving from each elevation is reconstructed by a weighted sum of received signals, which can be seen as a linear transform for the received signals. For distributed targets such as raindrops, however, the number of scatterers is significantly large, differently from the case of point targets that are standard targets in array signal processing. Thus, the spatial resolution of the conventional linear beamforming is limited. To improve the spatial resolution, we exploit two characteristics of a periodogram of each backscattered signal from the distributed targets. The periodogram is a series of the powers of the discrete Fourier transform (DFT) coefficients of each backscattered signal and utilized as a nonparametric estimate of the power spectral density . Since each power spectral density is proportional to the Doppler frequency distribution, 1) major components of the periodogram are concentrated in the vicinity of the mean Doppler frequency and 2) frequency indices of the major components are similar between adjacent elevations. These are expressed as group-sparsities of the DFT coefficient matrix of the backscattered signals, and we propose to reconstruct the signals through convex optimization exploiting the group-sparsities. We consider two optimization problems. One problem roughly evaluates the group-sparsities and is relatively easy to solve. The other evaluates the group-sparsities more accurately but requires more time to solve. Both problems are solved with the alternating direction method of multipliers (ADMM) including nonlinear mappings. Simulations using synthetic and real-world PAWR data show that the proposed method dramatically improves the spatial resolution.
- 1 January 2021
- Published by Acoustical Society of Japan in Acoustical Science and Technology
- Vol. 42 (6) , 305-313
WaveNet modeling of distortion pedal using spectral features
The present paper proposes a distortion pedal modeling method using the so-called WaveNet. A state-of-the-art method constructs a feedforward network by modifying the original autoregressive WaveNet, and trains it so that a loss function defined by the normalized mean squared error between the high-pass filtered outputs is minimized This method works well for pedals with low distortion, but not for those with high distortion. To solve this problem, the proposed method exploits the same WaveNet, but a novel loss function, which is defined by a weighted sum of errors in time and time-frequency (T-F) domains. The error in the time domain is defined by the mean squared error without the high-pass filtering, while that in the T-F domain is defined by a divergence between spectral features computed from the short-time Fourier transform. Numerical experiments using a pedal with high distortion, the Ibanez SD9, show that the proposed method is capable of precisely reproducing high-frequency components without attenuation of low-frequency components compared to the state-of-the-art method.
- 1 January 2021
- Published by Acoustical Society of Japan in Acoustical Science and Technology
- Vol. 42 (6) , 305-313
WaveNet modeling of distortion pedal using spectral features
The present paper proposes a distortion pedal modeling method using the so-called WaveNet. A state-of-the-art method constructs a feedforward network by modifying the original autoregressive WaveNet, and trains it so that a loss function defined by the normalized mean squared error between the high-pass filtered outputs is minimized This method works well for pedals with low distortion, but not for those with high distortion. To solve this problem, the proposed method exploits the same WaveNet, but a novel loss function, which is defined by a weighted sum of errors in time and time-frequency (T-F) domains. The error in the time domain is defined by the mean squared error without the high-pass filtering, while that in the T-F domain is defined by a divergence between spectral features computed from the short-time Fourier transform. Numerical experiments using a pedal with high distortion, the Ibanez SD9, show that the proposed method is capable of precisely reproducing high-frequency components without attenuation of low-frequency components compared to the state-of-the-art method.
- 16 October 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Signal Processing
- Vol. 66 (24) , 6363-6376
Exploiting Sparsity in Tight-Dimensional Spaces for Piecewise Continuous Signal Recovery
Recovery of certain piecewise continuous signals from noisy observations has been a major challenge in sciences and engineering. In this paper, in a tight-dimensional representation space, we exploit sparsity hidden in a class of possibly discontinuous signals named finite-dimensional piecewise continuous (FPC) signals. More precisely, we propose a tight-dimensional linear transformation which reveals a certain sparsity in discrete samples of the FPC signals. This transformation is designed by exploiting the fact that most of the consecutive samples are contained in special subspaces. Numerical experiments on recovery of piecewise polynomial signals and piecewise sinusoidal signals show the effectiveness of the revealed sparsity.
- 16 October 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Signal Processing
- Vol. 66 (24) , 6363-6376
Exploiting Sparsity in Tight-Dimensional Spaces for Piecewise Continuous Signal Recovery
Recovery of certain piecewise continuous signals from noisy observations has been a major challenge in sciences and engineering. In this paper, in a tight-dimensional representation space, we exploit sparsity hidden in a class of possibly discontinuous signals named finite-dimensional piecewise continuous (FPC) signals. More precisely, we propose a tight-dimensional linear transformation which reveals a certain sparsity in discrete samples of the FPC signals. This transformation is designed by exploiting the fact that most of the consecutive samples are contained in special subspaces. Numerical experiments on recovery of piecewise polynomial signals and piecewise sinusoidal signals show the effectiveness of the revealed sparsity.
- 1 July 2017
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 609-613
Recovery of piecewise finite-dimensional continuous signals by exploiting sparsity
Many inverse problems in science and engineering are formulated as recovery of piecewise finite-dimensional continuous (PFC) signals. Although the higher-order total variation (HTV) is known to be particularly effective for the sparsity-aware recovery of piecewise polynomials, it remains unclear so far whether the HTV can be extended to other signal models. In this paper, we present a convex regularizer which becomes a generalization of the HTV for the PFC signals. We first design a linear transformation which induces a certain group sparsity of samples of the PFC signals. This linear transformation is designed based on the fact that most of local samples can be interpolated by a fixed linear combination of known basis. Moreover, we provide theoretical evidence that the linear transformed samples have the group sparsity. Then, the proposed regularizer is designed to promote the group sparsity by using the ℓ 1,2 norm. A numerical experiment on recovery of piecewise sinusoidal signals shows the effectiveness of the proposed regularization.
- 1 July 2017
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 609-613
Recovery of piecewise finite-dimensional continuous signals by exploiting sparsity
Many inverse problems in science and engineering are formulated as recovery of piecewise finite-dimensional continuous (PFC) signals. Although the higher-order total variation (HTV) is known to be particularly effective for the sparsity-aware recovery of piecewise polynomials, it remains unclear so far whether the HTV can be extended to other signal models. In this paper, we present a convex regularizer which becomes a generalization of the HTV for the PFC signals. We first design a linear transformation which induces a certain group sparsity of samples of the PFC signals. This linear transformation is designed based on the fact that most of local samples can be interpolated by a fixed linear combination of known basis. Moreover, we provide theoretical evidence that the linear transformed samples have the group sparsity. Then, the proposed regularizer is designed to promote the group sparsity by using the ℓ 1,2 norm. A numerical experiment on recovery of piecewise sinusoidal signals shows the effectiveness of the proposed regularization.