arXiv subjects

Yuichi Tanaka

Publications and source records attributed to Yuichi Tanaka.

At least 19 recordsLinked to original sources

arXiv2026

WREN: Low Light Image Enhancement Using Retinex theory-based Double U-Net-like Structures

This paper proposes a neural network for low light image enhancement (LLIE) based on retinex theory to make LLIE robust for various dynamic range scenes. The retinex theory is an image formulation model inspired by a human color perception hypothesis, where a low light image is decomposed into intrinsic color context (i.e., reflectance map) and scene-dependent illumination (i.e., illumination map). Due to non-uniqueness of its decomposition, existing retinex-based LLIE methods often fail to achieve stable decomposition, which lead to over-enhancement. Typically, they are sensitive to the dynamic ranges that vary in different lighting conditions. To tackle this issue, we propose WREN: An LLIE neural network with double U-Net-like structures. WREN consists of two U-Net-like sub-networks. The first network has one encoder and two decoders that decompose an input image into the reflectance and illumination maps. The second network with a customized Transformer block between an encoder and a decoder only enhances the illumination map obtained from the first network: This completely follows the assumption of the retinex theory. Finally, the enhanced illumination map is recombined with the reflectance map. The network is trained end-to-end with a scale-invariant loss function, which gives robustness against the illumination scaling. Numerical results show that our method achieves the state-of-the-art performance across multiple datasets. Our code is available online.

cs.CV↗

arXiv2026

Lossy compression of weighted graph adjacency matrices by transform coding

In this paper, we propose a compression framework for weighted graphs in which the graph topology is transmitted losslessly and edge weights are compressed lossily. A challenge in the lossy compression of edge weights is that the underlying relationships between edges are ambiguous. To address this issue, we first transform the unweighted graph into the corresponding line graph, whose nodes represent the edges of the original graph and whose edges encode the relationships between them. The line graph transform allows us to regard edge weights as a graph signal defined on the line graph. Instead of transmitting the edge-weight vector, we first transform it with a graph filter bank on the line graph. Then, quantization and entropy coding are performed on the transformed coefficients of the edge weight vector. In addition to the lossy compression method, we formalize edge smoothness on the line graph and show that it serves as a measure of the difficulty of compression. The proposed smoothness measure can be easily calculated without converting to a line graph. This provides insight into the expected compression performance of a given weighted graph. Experiments on synthetic and real-world data validate the effectiveness of the proposed method by comparing it with existing matrix preprocessing methods.

cs.IT↗

arXiv2026

Multimodal Signal Restoration with Signed Twofold Graph Learning

Multimodal signals on sensor networks are commonly modeled under the twofold graph assumption (TGA), which represents spatial structure and inter-modality relations as two separate graphs. Existing TGA-based signal restoration methods, however, either assume the graphs are known or restrict edge weights to be non-negative, preventing them from capturing negative inter-modal correlations. We address both limitations by formulating joint signal restoration and twofold graph learning as MAP estimation under a matrix normal prior, where the spatial and modality graph Laplacians appear directly as precision matrices. The resulting non-convex objective is solved by alternating minimization: The signal is updated via conjugate gradient applied to the arising Sylvester-type linear system; the graphs are updated via primal-dual hybrid gradient (PDHG). We further propose a method to estimate the signed structure of the modality graph from the dominant eigenspace of a complementary kernel matrix, which is then used in PDHG to update edge magnitudes. These iterative solvers are then unrolled into a feedforward network, with regularization weights and step sizes as layer-wise trainable parameters. Experiments on synthetic multimodal graph signals and a real Japan meteorological dataset confirm that the proposed method outperforms existing baselines across a range of noise levels and missing-data patterns.

eess.SP↗

arXiv2026

Joint Alignment and Denoising for Event-Based Vision Sensors Using Regret-based Pareto Optimization

This paper proposes a joint alignment and denoising method for event-based vision sensors (EVSs). Existing signal processing methods for EVSs typically perform event alignment (EA) and event denoising (ED) as separate modules. However, this separation creates a dilemma: without ED, EA is biased by noise, whereas without EA, ED struggles to distinguish signal events from noise ones. To address this dilemma, we jointly optimize EA and ED by formulating a bi-objective Pareto optimization problem. Our formulation is built upon a contrast map that counts the number of events localized in each pixel. With a contrast map, we can formulate EA as maximizing its variance and ED as minimizing the variance. We cast these two conflicting problems as a Pareto optimization and use a regret strategy to obtain a solution. Experimental results on denoising and motion estimation demonstrate that our method achieves improvements against alternative ones.

eess.IV↗

arXiv2026

Denoising for Neuromorphic Cameras Based on Graph Spectral Features

Neuromorphic cameras, also known as event-based cameras, can detect changes in the environmental brightness asynchronously and independently for each pixel. They output the brightness changes, i.e., events, as 3-D (2-D pixel coordinates + time) streaming data. While event-based cameras are used in many applications because of their desirable characteristics, e.g., high temporal resolution, low latency, low power consumption, and high dynamic range, their measurements contain considerable noise due to their high sensitivity. In this paper, we propose a denoising method for event-based cameras based on graph spectral features. In the proposed method, we first construct a graph where nodes represent events and edges represent the spatiotemporal distance between the events. To calculate the graph-specified parameter that controls the connectivities of a constructed graph, we utilize the prior on the density of 3-D events. We then calculate the eigenvectors of the graph Laplacian. The obtained eigenvectors are used to extract noiseless events directly. In the calculation of the eigenvectors, we customize the graph Laplacian to reorder its eigenvalues. This allows us to leverage fast eigensolver algorithms instead of the naive eigendecomposition and thereby reduce computational complexity. In experiments on synthetic and real-world event data, we demonstrate that the proposed method effectively removes noise events from the raw events compared to alternative methods.

eess.IV↗

arXiv2026

Dynamic Sensor Scheduling Based on Node Partitioning of Graphs

This paper proposes a dynamic sensor scheduling method for sensor networks. In sensor network applications, we often need multiple equally-informative node subsets that are activated sequentially to make a sensor network robust against concentrated battery consumption and sensor failures. In addition, quality of these subsets changes dynamically and thus we must adapt those changes. To find those node subsets, we propose a graph node partitioning method based on sampling theory for graph signals. We aim to minimize the average reconstruction error for signals obtained at all node subsets, in contrast to conventional single subset selection. The graph node partitioning problem is formulated as a difference-of-convex (DC) optimization based on a subspace prior of graph signals, and is solved by the proximal DC algorithm. It guarantees convergence to a critical point. To accommodate the online scenario where the signal subspace and optimal partitioning may change over time, we adaptively estimate the signal subspace from historical data and sequentially update the prior for our partitioning method. Numerical experiments on synthetic and real-world sensor network data demonstrate that the proposed method achieves lower average mean squared errors compared to alternative methods.

eess.SP↗

arXiv2026

Auditory Attention Decoding without Spatial Information: A Diotic EEG Study

Auditory attention decoding (AAD) identifies the attended speech stream in multi-speaker environments by decoding brain signals such as electroencephalography (EEG). This technology is essential for realizing smart hearing aids that address the cocktail party problem and for facilitating objective audiometry systems. Existing AAD research mainly utilizes dichotic environments where different speech signals are presented to the left and right ears, enabling models to classify directional attention rather than speech content. However, this spatial reliance limits applicability to real-world scenarios, such as the "cocktail party" situation, where speakers overlap or move dynamically. To address this challenge, we propose an AAD framework for diotic environments where identical speech mixtures are presented to both ears, eliminating spatial cues. Our approach maps EEG and speech signals into a shared latent space using independent encoders. We extract speech features using wav2vec 2.0 and encode them with a 2-layer 1D convolutional neural network (CNN), while employing the BrainNetwork architecture for EEG encoding. The model identifies the attended speech by calculating the cosine similarity between EEG and speech representations. We evaluate our method on a diotic EEG dataset and achieve 72.70% accuracy, which is 22.58% higher than the state-of-the-art direction-based AAD method.

eess.SP↗

arXiv2025

Graph Signal Denoising Using Regularization by Denoising and Its Parameter Estimation

In this paper, we propose an interpretable denoising method for graph signals using regularization by denoising (RED). RED is a technique developed for image restoration that uses an efficient (and sometimes black-box) denoiser in the regularization term of the optimization problem. By using RED, optimization problems can be designed with the explicit use of the denoiser, and the gradient of the regularization term can be easily computed under mild conditions. We adapt RED for denoising of graph signals beyond image processing. We show that many graph signal denoisers, including graph neural networks, theoretically or practically satisfy the conditions for RED. We also study the effectiveness of RED from a graph filter perspective. Furthermore, we propose supervised and unsupervised parameter estimation methods based on deep algorithm unrolling. These methods aim to enhance the algorithm applicability, particularly in the unsupervised setting. Denoising experiments for synthetic and real-world datasets show that our proposed method improves signal denoising accuracy in mean squared error compared to existing graph signal denoising methods.

eess.SP↗

arXiv2025

Efficient Learning of Balanced Signed Graphs via Sparse Linear Programming

Signed graphs are equipped with both positive and negative edge weights, encoding pairwise correlations as well as anti-correlations in data. A balanced signed graph is a signed graph with no cycles containing an odd number of negative edges. Laplacian of a balanced signed graph has eigenvectors that map via a simple linear transform to ones in a corresponding positive graph Laplacian, thus enabling reuse of spectral filtering tools designed for positive graphs. We propose an efficient computation method to learn a balanced signed graph Laplacian directly from data. Specifically, extending a previous linear programming (LP) based sparse inverse covariance estimation method called CLIME, we formulate a new LP problem for each Laplacian column $i$, where the linear constraints restrict weight signs of edges stemming from node $i$, so that nodes of same / different polarities are connected by positive / negative edges. We derive a feasible CLIME parameter $\rho_i$ for each sign-constrained column problem. We solve the LP problem efficiently by tailoring a sparse LP method based on ADMM. We theoretically prove that the row / column updates produce a non-increasing objective sequence, and show that the iterations are terminated in a finite number of steps. Extensive experimental results on synthetic and real-world datasets show that our balanced graph learning method outperforms competing methods and enables reuse of spectral filters, wavelets, and graph neural nets (GNN) constructed for positive graphs.

cs.LG↗

arXiv2025

Algorithm Unrolling-based Denoising of Multimodal Graph Signals

We propose a denoising method for multimodal graph signals by an alternating minimization scheme that sequentially solves signal restoration and graph learning problems. Many complex-structured data, i.e., those on sensor networks, can capture multiple modalities at each measurement point, referred to as modalities. They are also assumed to have an underlying structure or correlations in modality as well as space. Such multimodal data are regarded as graph signals on a twofold graph and they are often corrupted by noise. Furthermore, their spatial/modality relationships are not always given a priori: We need to estimate twofold graphs during a denoising algorithm. In this paper, we consider a signal denoising method on twofold graphs, where graphs are learned simultaneously. Specifically, the graph learning subproblems are solved using the primal-dual splitting (PDS) algorithm, while the signal update has a closed-form solution. Parameters in this iterative algorithm are learned from training data by unrolling the iteration with deep algorithm unrolling. Experimental results on synthetic and real-world data demonstrate that the proposed method outperforms existing model- and deep learning-based graph signal denoising methods.

eess.SP↗

arXiv2025

Joint Graph Estimation and Signal Restoration for Robust Federated Learning

We propose a robust aggregation method for model parameters in federated learning (FL) under noisy communications. FL is a distributed machine learning paradigm in which a central server aggregates local model parameters from multiple clients. These parameters are often noisy and/or have missing values during data collection, training, and communication between the clients and server. This may cause a considerable drop in model accuracy. To address this issue, we learn a graph that represents pairwise relationships between model parameters of the clients during aggregation. We realize it with a joint problem of graph learning and signal (i.e., model parameters) restoration. The problem is formulated as a difference-of-convex (DC) optimization, which is efficiently solved via a proximal DC algorithm. Experimental results on MNIST and CIFAR-10 datasets show that the proposed method outperforms existing approaches by up to $2$--$5\%$ in classification accuracy under biased data distributions and noisy conditions.

cs.LG↗

arXiv2024

Multiscale Graph Construction Using Non-local Cluster Features

This paper presents a multiscale graph construction method using both graph and signal features. Multiscale graph is a hierarchical representation of the graph, where a node at each level indicates a cluster in a finer resolution. To obtain the hierarchical clusters, existing methods often use graph clustering; however, they may ignore signal variations. As a result, these methods could fail to detect the clusters having similar features on nodes. In this paper, we consider graph and node-wise features simultaneously for multiscale clustering of a graph. With given clusters of the graph, the clusters are merged hierarchically in three steps: 1) Feature vectors in the clusters are extracted. 2) Similarities among cluster features are calculated using optimal transport. 3) A variable $k$-nearest neighbor graph (V$k$NNG) is constructed and graph spectral clustering is applied to the V$k$NNG to obtain clusters at a coarser scale. Additionally, the multiscale graph in this paper has \textit{non-local} characteristics: Nodes with similar features are merged even if they are spatially separated. In experiments on multiscale image and point cloud segmentation, we demonstrate the effectiveness of the proposed method.

cs.CV↗

arXiv2024

Time-Varying Graph Signal Estimation among Multiple Sub-Networks

This paper presents an estimation method for time-varying graph signals among multiple sub-networks. In many sensor networks, signals observed are associated with nodes (i.e., sensors), and edges of the network represent the inter-node connectivity. For a large sensor network, measuring signal values at all nodes over time requires huge resources, particularly in terms of energy consumption. To alleviate the issue, we consider a scenario that a sub-network, i.e., cluster, from the whole network is extracted and an intra-cluster analysis is performed based on the statistics in the cluster. The statistics are then utilized to estimate signal values in another cluster. This leads to the requirement for transferring a set of parameters of the sub-network to the others, while the numbers of nodes in the clusters are typically different. In this paper, we propose a cooperative Kalman filter between two sub-networks. The proposed method alternately estimates signals in time between two sub-networks. We formulate a state-space model in the source cluster and transfer it to the target cluster on the basis of optimal transport. In the signal estimation experiments of synthetic and real-world signals, we validate the effectiveness of the proposed method.

eess.SP↗

arXiv2024

Efficient Learning of Balanced Signed Graphs via Iterative Linear Programming

Signed graphs are equipped with both positive and negative edge weights, encoding pairwise correlations as well as anti-correlations in data. A balanced signed graph has no cycles of odd number of negative edges. Laplacian of a balanced signed graph has eigenvectors that map simply to ones in a similarity-transformed positive graph Laplacian, thus enabling reuse of well-studied spectral filters designed for positive graphs. We propose a fast method to learn a balanced signed graph Laplacian directly from data. Specifically, for each node $i$, to determine its polarity $\beta_i \in \{-1,1\}$ and edge weights $\{w_{i,j}\}_{j=1}^N$, we extend a sparse inverse covariance formulation based on linear programming (LP) called CLIME, by adding linear constraints to enforce ``consistent" signs of edge weights $\{w_{i,j}\}_{j=1}^N$ with the polarities of connected nodes -- i.e., positive/negative edges connect nodes of same/opposing polarities. For each LP, we adapt projections on convex set (POCS) to determine a suitable CLIME parameter $\rho > 0$ that guarantees LP feasibility. We solve the resulting LP via an off-the-shelf LP solver in $\mathcal{O}(N^{2.055})$. Experiments on synthetic and real-world datasets show that our balanced graph learning method outperforms competing methods and enables the use of spectral filters and graph convolutional networks (GCNs) designed for positive graphs on signed graphs.

cs.LG↗

arXiv2024

Constructing an Interpretable Deep Denoiser by Unrolling Graph Laplacian Regularizer

An image denoiser can be used for a wide range of restoration problems via the Plug-and-Play (PnP) architecture. In this paper, we propose a general framework to build an interpretable graph-based deep denoiser (GDD) by unrolling a solution to a maximum a posteriori (MAP) problem equipped with a graph Laplacian regularizer (GLR) as signal prior. Leveraging a recent theorem showing that any (pseudo-)linear denoiser $\boldsymbol \Psi$, under mild conditions, can be mapped to a solution of a MAP denoising problem regularized using GLR, we first initialize a graph Laplacian matrix $\mathbf L$ via truncated Taylor Series Expansion (TSE) of $\boldsymbol \Psi^{-1}$. Then, we compute the MAP linear system solution by unrolling iterations of the conjugate gradient (CG) algorithm into a sequence of neural layers as a feed-forward network -- one that is amenable to parameter tuning. The resulting GDD network is "graph-interpretable", low in parameter count, and easy to initialize thanks to $\mathbf L$ derived from a known well-performing denoiser $\boldsymbol \Psi$. Experimental results show that GDD achieves competitive image denoising performance compared to competitors, but employing far fewer parameters, and is more robust to covariate shift.

eess.IV↗

arXiv2024

Edge Sampling of Graphs: Graph Signal Processing Approach With Edge Smoothness

Finding important edges in a graph is a crucial problem for various research fields, such as network epidemics, signal processing, machine learning, and sensor networks. In this paper, we tackle the problem based on sampling theory on graphs. We convert the original graph to a line graph where its nodes and edges, respectively, represent the original edges and the connections between the edges. We then perform node sampling of the line graph based on the edge smoothness assumption: This process selects the most critical edges in the original graph. We present a general framework of edge sampling based on graph sampling theory and reveal a theoretical relationship between the degree of the original graph and the line graph. We also propose an acceleration method for edge sampling in the proposed framework by using the relationship between two types of Laplacian of the node and edge domains. Experimental results in synthetic and real-world graphs validate the effectiveness of our approach against some alternative edge selection methods.

eess.SP↗

arXiv2024

PHISWID: Physics-Inspired Underwater Image Dataset Synthesized from RGB-D Images

This paper introduces the physics-inspired synthesized underwater image dataset (PHISWID), a dataset tailored for enhancing underwater image processing through physics-inspired image synthesis. For underwater image enhancement, data-driven approaches (e.g., deep neural networks) typically demand extensive datasets, yet acquiring paired clean atmospheric images and degraded underwater images poses significant challenges. Existing datasets have limited contributions to image enhancement due to lack of physics models, publicity, and ground-truth atmospheric images. PHISWID addresses these issues by offering a set of paired atmospheric and underwater images. Specifically, underwater images are synthetically degraded by color degradation and marine snow artifacts from atmospheric RGB-D images. It is enabled based on a physics-based underwater image observation model. Our synthetic approach generates a large quantity of the pairs, enabling effective training of deep neural networks and objective image quality assessment. Through benchmark experiments with some datasets and image enhancement methods, we validate that our dataset can improve the image enhancement performance. Our dataset, which is publicly available, contributes to the development in underwater image processing.

cs.CV↗

arXiv2024

Lossy Compression of Adjacency Matrices by Graph Filter Banks

This paper proposes a compression framework for adjacency matrices of weighted graphs based on graph filter banks. Adjacency matrices are widely used mathematical representations of graphs and are used in various applications in signal processing, machine learning, and data mining. In many problems of interest, these adjacency matrices can be large, so efficient compression methods are crucial. In this paper, we propose a lossy compression of weighted adjacency matrices, where the binary adjacency information is encoded losslessly (so the topological information of the graph is preserved) while the edge weights are compressed lossily. For the edge weight compression, the target graph is converted into a line graph, whose nodes correspond to the edges of the original graph, and where the original edge weights are regarded as a graph signal on the line graph. We then transform the edge weights on the line graph with a graph filter bank for sparse representation. Experiments on synthetic data validate the effectiveness of the proposed method by comparing it with existing lossy matrix compression methods.

eess.SP↗