arXiv subjects

Sanjay Ghosh

Publications and source records attributed to Sanjay Ghosh.

At least 19 recordsLinked to original sources

arXiv2026

LUMEN: Low-light Unified Multi-stage Enhancement Network using depth-guided flash, clustering, and attention-based Transformers

Low-light image enhancement remains a challenging problem due to severe noise, color distortion, contrast degradation, and loss of structural details under insufficient illumination. Existing methods typically apply uniform enhancement without considering the depth-dependent nature of light attenuation and sensor noise in real-world scenes. To address this limitation, we propose LUMEN, a multi-stage enhancement framework that integrates virtual flash simulation with transformer-based feature fusion. The proposed framework first estimates scene depth from low-light inputs using a dedicated encoder-decoder network, after which a soft clustering module partitions pixels into depth-aware regions, enabling depth-dependent flash simulation. The simulated flash features, together with depth representations, are fused with image features through efficient attention-based fusion blocks to enhance global context while preserving fine details. A composite loss function combining reconstruction, perceptual, structural, color, edge, and depth consistency objectives ensures both visual fidelity and perceptual quality. Extensive experiments on LOL-v1 and LOL-v2 benchmarks demonstrate that LUMEN achieves state-of-the-art performance and produces visually natural results compared with several state-of-the-art methods.

eess.IV↗

arXiv2026

An Attention-Enhanced Network with Joint Dehazing and Retinex-Based Enhancement for Underwater Images

Underwater images suffer from severe wavelength-dependent light absorption and scattering, and turbidity due to suspended particles, degrading visual quality for applications in autonomous underwater vehicles (AUVs), marine biology, archaeology, and offshore infrastructure inspection. Classical IFM inadequately capture nonlinear underwater light behavior, while purely data-driven methods lack physical interpretability. This paper proposes a three-stage network named ADR, that extends the underwater image formation model with additional terms to perform underwater dehazing, followed by Retinex-based enhancement and attention-enabled U-Net++ refinement. Experiments on UIEB and UFO-120 benchmark datasets demonstrate competitive performance with state-of-the-art methods.

eess.IV↗

arXiv2026

An Underwater Dehazing Network with Implicit Transmission Estimation

Underwater images suffer from wavelength-dependent light absorption and scattering, which reduces visual quality. This phenomenon could limit the operational reliability of autonomous underwater vehicles, marine surveys, and offshore inspection systems. Purely classical methods often achieve suboptimal performance in real-world datasets, while purely data-driven methods lack physical interpretability. In this letter, we propose UDehaze-iT, a deep network for underwater image enhancement that estimates scene depth implicitly and derives per-channel transmission through the Beer-Lambert law with learnable attenuation coefficients. We estimate atmospheric light as a semi-classical per-channel scalar, and a zero-initialized residual refiner corrects remaining artefacts after dehazing. To effectively train our method, we apply a composite loss function consisting of five key terms: a L1 loss, a multi-scale patchwise DCT loss, a forward model reconstruction loss, and two regularization terms. With ~0.9M parameters, UDehaze-iT achieves competitive performance on UIEB and UFO-120 datasets.

eess.IV↗

arXiv2025

Learned Adaptive Kernels for High-Fidelity Image Downscaling

Image downscaling is a fundamental operation in image processing, crucial for adapting high-resolution content to various display and storage constraints. While classic methods often introduce blurring or aliasing, recent learning-based approaches offer improved adaptivity. However, achieving maximal fidelity against ground-truth low-resolution (LR) images, particularly by accounting for channel-specific characteristics, remains an open challenge. This paper introduces ADK-Net (Adaptive Downscaling Kernel Network), a novel deep convolutional neural network framework for high-fidelity supervised image downscaling. ADK-Net explicitly addresses channel interdependencies by learning to predict spatially-varying, adaptive resampling kernels independently for each pixel and uniquely for each color channel (RGB). The architecture employs a hierarchical design featuring a ResNet-based feature extractor and parallel channel-specific kernel generators, themselves composed of ResNet-based trunk and branch sub-modules, enabling fine-grained kernel prediction. Trained end-to-end using an L1 reconstruction loss against ground-truth LR data, ADK-Net effectively learns the target downscaling transformation. Extensive quantitative and qualitative experiments on standard benchmarks, including the RealSR dataset, demonstrate that ADK-Net establishes a new state-of-the-art in supervised image downscaling, yielding significant improvements in PSNR and SSIM metrics compared to existing learning-based and traditional methods.

eess.IV↗

arXiv2025

Functional Connectome Fingerprinting Using Convolutional and Dictionary Learning

Advances in data analysis and machine learning have revolutionized the study of brain signatures using fMRI, enabling non-invasive exploration of cognition and behavior through individual neural patterns. Functional connectivity (FC), which quantifies statistical relationships between brain regions, has emerged as a key metric for studying individual variability and developing biomarkers for personalized medicine in neurological and psychiatric disorders. The concept of subject fingerprinting, introduced by Finn et al. (2015), leverages neural connectivity variability to identify individuals based on their unique patterns. While traditional FC methods perform well on small datasets, machine learning techniques are more effective with larger datasets, isolating individual-specific features and maximizing inter-subject differences. In this study, we propose a framework combining convolutional autoencoders and sparse dictionary learning to enhance fingerprint accuracy. Autoencoders capture shared connectivity patterns while isolating subject-specific features in residual FC matrices, which are analyzed using sparse coding to identify distinctive features. Tested on the Human Connectome Project dataset, this approach achieved a 10% improvement over baseline group-averaged FC models. Our results highlight the potential of integrating deep learning and sparse coding techniques for scalable and robust functional connectome fingerprinting, advancing personalized neuroscience applications and biomarker development.

eess.IV↗

arXiv2025

Graph-Guided Structured Recovery of Modulo EEG Signals

Electroencephalography (EEG) acquisition is often affected by large-amplitude variations across subjects and sessions. This makes conventional front ends vulnerable to saturation and clipping. Modulo sampling addresses this limitation by folding the signal into a bounded range rather than truncating it. The recovery task is then to reconstruct the original waveform from periodic observations, which is a highly ill-posed inverse problem. In this work, we propose a compact signal-structured framework for modulo EEG recovery. The method combines phase-aware features, graph-temporal modeling, and structured fold-state decoding. A boundary-guided gating technique improves inference near wrap events, where uncertainty is highest. Experiments across two EEG datasets, over a wide range of folding thresholds, demonstrate consistent recovery performance. Ablation and noise studies further support the contribution of the structured decoding technique. The proposed signal-processing approach effectively recovers folded EEG recordings. These results show that structured fold-state decoding is effective for recovering folded EEG. The code is accessible at https://github.com/soujo/graphunwrapnet.

eess.SP↗

arXiv2025

Selection and Stability of Functional Connectivity Features for Classification of Brain Disorders

Brain disorders are an umbrella term for a group of neurological and psychiatric conditions that have a major effect on thinking, feeling, and acting. These conditions encompass a wide range of conditions. The illnesses in question pose significant difficulties not only for individuals, but also for healthcare systems all across the world. In this study, we explore the capability of explainable machine learning for classification of people who suffer from brain disorders. This is accomplished by the utilization of brain connection map, also referred as connectome, derived from functional magnetic resonance imaging (fMRI) data. In order to analyze features that are based on the connectome, we investigated several different feature selection procedures. These strategies included the Least Absolute Shrinkage and Selection Operator (LASSO), Relief, and Analysis of Variance (ANOVA), in addition to a logistic regression (LR) classifier. First and foremost, the purpose was to evaluate and contrast the classification accuracy of different feature selection methods in terms of distinguishing healthy controls from diseased individuals. The evaluation of the stability of the traits that were chosen was the second objective. The identification of the regions of the brain that have an effect on the classification was the third main objective. When applied to the UCLA dataset, the LASSO approach, which is our most effective strategy, produced a classification accuracy of 91.85% and a stability index of 0.74, which is greater than the results obtained by other approaches: Relief and ANOVA. These methods are effective in locating trustworthy biomarkers, which adds to the development of connectome-based classification in the context of issues that impact the brain.

q-bio.NC↗

arXiv2025

Bridging Accuracy and Explainability in EEG-based Graph Attention Network for Depression Detection

Depression is a major cause of global mental illness and significantly influences suicide rates. Timely and accurate diagnosis is essential for effective intervention. Electroencephalography (EEG) provides a non-invasive and accessible method for examining cerebral activity and identifying disease-associated patterns. We propose a novel graph-based deep learning framework, named Edge-gated, axis-mixed Pooling Attention Network (ExPANet), for differentiating major depressive disorder (MDD) patients from healthy controls (HC). EEG recordings undergo preprocessing to eliminate artifacts and are segmented into short periods of activity. We extract 14 features from each segment, which include time, frequency, fractal, and complexity domains. Electrodes are represented as nodes, whereas edges are determined by the phase-locking value (PLV) to represent functional connectivity. The generated brain graphs are examined utilizing an adapted graph attention network. This architecture acquires both localized electrode characteristics and comprehensive functional connectivity patterns. The proposed framework attains superior performance relative to current EEG-based approaches across two different datasets. A fundamental advantage of our methodology is its explainability. We evaluated the significance of features, channels, and edges, in addition to intrinsic attention weights. These studies highlight features, cerebral areas, and connectivity associations that are especially relevant to MDD, many of which correspond with clinical data. Our findings demonstrate a reliable and transparent method for EEG-based screening of MDD, using deep learning with clinically relevant results.

eess.SP↗

arXiv2025

High-Quality and Large-Scale Image Downscaling for Modern Display Devices

In modern display technology and visualization tools, downscaling images is one of the most important activities. This procedure aims to maintain both visual authenticity and structural integrity while reducing the dimensions of an image at a large scale to fit the dimension of the display devices. In this study, we proposed a new technique for downscaling images that uses co-occurrence learning to maintain structural and perceptual information while reducing resolution. The technique uses the input image to create a data-driven co-occurrence profile that captures the frequency of intensity correlations in nearby neighborhoods. A refined filtering process is guided by this profile, which acts as a content-adaptive range kernel. The contribution of each input pixel is based on how closely it resembles pair-wise intensity values with it's neighbors. We validate our proposed technique on four datasets: DIV2K, BSD100, Urban100, and RealSR to show its effective downscaling capacity. Our technique could obtain up to 39.22 dB PSNR on the DIV2K dataset and PIQE up to 26.35 on the same dataset when downscaling by 8x and 16x, respectively. Numerous experimental findings attest to the ability of the suggested picture downscaling method to outperform more contemporary approaches in terms of both visual quality and performance measures. Unlike most existing methods, which did not focus on the large-scale image resizing scenario, we achieve high-quality downscaled images without texture loss or edge blurring. Our method, LSID (large scale image downscaling), successfully preserves high-frequency structures like edges, textures, and repeating patterns by focusing on statistically consistent pixels while reducing aliasing and blurring artifacts that are typical of traditional downscaling techniques.

eess.IV↗

arXiv2025

Low-Light Image Enhancement Using Gamma Learning And Attention-Enabled Encoder-Decoder Networks

Images acquired in low-light environments present significant obstacles for computer vision systems and human perception, especially for applications requiring accurate object recognition and scene analysis. Such images typically manifest multiple quality issues: amplified noise, inadequate scene illumination, contrast reduction, color distortion, and loss of details. While recent deep learning methods have shown promise, developing simple and efficient frameworks that naturally integrate global illumination adjustment with local detail refinement continues to be an important objective. To this end, we introduce a dual-stage deep learning architecture that combines adaptive gamma correction with attention-enhanced refinement to address these fundamental limitations. The first stage uses an Adaptive Gamma Correction Module (AGCM) to learn suitable gamma values for each pixel based on both local and global cues, producing a brightened intermediate output. The second stage applies an encoder-decoder deep network with Convolutional Block Attention Modules (CBAM) to this brightened image, in order to restore finer details. We train the network using a composite loss that includes L1 reconstruction, SSIM, total variation, color constancy, and gamma regularization terms to balance pixel accuracy with visual quality. Experiments on LOL-v1, LOL-v2 real, and LOL-v2 synthetic datasets show our method reaches PSNR of upto 29.96 dB and upto 0.9458 SSIM, outperforming existing approaches. Additional tests on DICM, LIME, MEF, and NPE datasets using NIQE, BRISQUE, and UNIQUE metrics confirm better perceptual quality with fewer artifacts, achieving the best NIQE scores across all datasets. Our GAtED (Gamma learned and Attention-enabled Encoder-Decoder) method effectively handles both global illumination adjustment and local detail enhancement, offering a practical solution for low-light enhancement.

eess.IV↗

arXiv2025

Structure Aware Image Downscaling

Image downscaling is one of the key operations in recent display technology and visualization tools. By this process, the dimension of an image is reduced, aiming to preserve structural integrity and visual fidelity. In this paper, we propose a new image downscaling method which is built on the core ideas of image filtering and edge detection. In particular, we present a structure-informed downscaling algorithm that maintains fine details through edge-aware processing. The proposed method comprises three steps: (i) edge map computation, (ii) edge-guided interpolation, and (iii) texture enhancement. To faithfully retain the strong structures in an image, we first compute the edge maps by applying an efficient edge detection operator. This is followed by an edge-guided interpolation to preserve fine details after resizing. Finally, we fuse local texture enriched component of the original image to the interpolated one to restore high-frequency information. By integrating edge information with adaptive filtering, our approach effectively minimizes artifacts while retaining crucial image features. To demonstrate the effective downscaling capability of our proposed method, we validate on four datasets: DIV2K, BSD100, Urban100, and RealSR. For downscaling by 4x, our method could achieve as high as 39.07 dB PSNR on the DIV2K dataset and 38.71 dB on the RealSR dataset. Extensive experimental results confirm that the proposed image downscaling method is capable of achieving superior performance in terms of both visual quality and performance metrics with reference to recent methods. Most importantly, the downscaled images by our method do not suffer from edge blurring and texture loss, unlike many existing ones.

eess.IV↗

arXiv2025

Poisson Informed Retinex Network for Extreme Low-Light Image Enhancement

Low-light image denoising and enhancement are challenging, especially when traditional noise assumptions, such as Gaussian noise, do not hold in majority. In many real-world scenarios, such as low-light imaging, noise is signal-dependent and is better represented as Poisson noise. In this work, we address the problem of denoising images degraded by Poisson noise under extreme low-light conditions. We introduce a light-weight deep learning-based method that integrates Retinex based decomposition with Poisson denoising into a unified encoder-decoder network. The model simultaneously enhances illumination and suppresses noise by incorporating a Poisson denoising loss to address signal-dependent noise. Without prior requirement for reflectance and illumination, the network learns an effective decomposition process while ensuring consistent reflectance and smooth illumination without causing any form of color distortion. The experimental results demonstrate the effectiveness and practicality of the proposed low-light illumination enhancement method. Our method significantly improves visibility and brightness in low-light conditions, while preserving image structure and color constancy under ambient illumination.

eess.IV↗

arXiv2021

Efficient Hierarchical Bayesian Inference for Spatio-temporal Regression Models in Neuroimaging

Several problems in neuroimaging and beyond require inference on the parameters of multi-task sparse hierarchical regression models. Examples include M/EEG inverse problems, neural encoding models for task-based fMRI analyses, and climate science. In these domains, both the model parameters to be inferred and the measurement noise may exhibit a complex spatio-temporal structure. Existing work either neglects the temporal structure or leads to computationally demanding inference schemes. Overcoming these limitations, we devise a novel flexible hierarchical Bayesian framework within which the spatio-temporal dynamics of model parameters and noise are modeled to have Kronecker product covariance structure. Inference in our framework is based on majorization-minimization optimization and has guaranteed convergence properties. Our highly efficient algorithms exploit the intrinsic Riemannian geometry of temporal autocovariance matrices. For stationary dynamics described by Toeplitz matrices, the theory of circulant embeddings is employed. We prove convex bounding properties and derive update rules of the resulting algorithms. On both synthetic and real neural data from M/EEG, we demonstrate that our methods lead to improved performance.

stat.ML↗

arXiv2020

Nonlocal Co-occurrence for Image Downscaling

Image downscaling is one of the widely used operations in image processing and computer graphics. It was recently demonstrated in the literature that kernel-based convolutional filters could be modified to develop efficient image downscaling algorithms. In this work, we present a new downscaling technique which is based on kernel-based image filtering concept. We propose to use pairwise co-occurrence similarity of the pixelpairs as the range kernel similarity in the filtering operation. The co-occurrence of the pixel-pair is learned directly from the input image. This co-occurrence learning is performed in a neighborhood based fashion all over the image. The proposed method can preserve the high-frequency structures, which were present in the input image, into the downscaled image. The idea is further extended to the case of fractions factor of downscaling. The resulting images retain visually-important details and do not suffer from edge-blurring artifact. We demonstrate the effectiveness of our proposed approach with extensive experiments on a large number of images downscaled with various downscaling factors.

eess.IV↗

arXiv2020

Model-Based Deep Learning for Reconstruction of Joint k-q Under-sampled High Resolution Diffusion MRI

We propose a model-based deep learning architecture for the reconstruction of highly accelerated diffusion magnetic resonance imaging (MRI) that enables high resolution imaging. The proposed reconstruction jointly recovers all the diffusion weighted images in a single step from a joint k-q under-sampled acquisition in a parallel MRI setting. We propose the novel use of a pre-trained denoiser as a regularizer in a model-based reconstruction for the recovery of highly under-sampled data. Specifically, we designed the denoiser based on a general diffusion MRI tissue microstructure model for multi-compartmental modeling. By using a wide range of biologically plausible parameter values for the multi-compartmental microstructure model, we simulated diffusion signal that spans the entire microstructure parameter space. A neural network was trained in an unsupervised manner using an autoencoder to learn the diffusion MRI signal subspace. We employed the autoencoder in a model-based reconstruction and show that the autoencoder provides a strong denoising prior to recover the q-space signal. We show reconstruction results on a simulated brain dataset that shows high acceleration capabilities of the proposed method.

eess.IV↗

arXiv2019

Linearized ADMM and Fast Nonlocal Denoising for Efficient Plug-and-Play Restoration

In plug-and-play image restoration, the regularization is performed using powerful denoisers such as nonlocal means (NLM) or BM3D. This is done within the framework of alternating direction method of multipliers (ADMM), where the regularization step is formally replaced by an off-the-shelf denoiser. Each plug-and-play iteration involves the inversion of the forward model followed by a denoising step. In this paper, we present a couple of ideas for improving the efficiency of the inversion and denoising steps. First, we propose to use linearized ADMM, which generally allows us to perform the inversion at a lower cost than standard ADMM. Moreover, we can easily incorporate hard constraints into the optimization framework as a result. Second, we develop a fast algorithm for doubly stochastic NLM, originally proposed by Sreehari et al. (IEEE TCI, 2016), which is about 80x faster than brute-force computation. This particular denoiser can be expressed as the proximal map of a convex regularizer and, as a consequence, we can guarantee convergence for linearized plug-and-play ADMM. We demonstrate the effectiveness of our proposals for super-resolution and single-photon imaging.

cs.CV↗

arXiv2018

Optimized Fourier Bilateral Filtering

We consider the problem of approximating a truncated Gaussian kernel using Fourier (trigonometric) functions. The computation-intensive bilateral filter can be expressed using fast convolutions by applying such an approximation to its range kernel, where the truncation in question is the dynamic range of the input image. The error from such an approximation depends on the period, the number of sinusoids, and the coefficient of each sinusoid. For a fixed period, we recently proposed a model for optimizing the coefficients using least-squares fitting. Following the Compressive Bilateral Filter (CBF), we demonstrate that the approximation can be improved by taking the period into account during the optimization. The accuracy of the resulting filtering is found to be at least as good as CBF, but significantly better for certain cases. The proposed approximation can also be used for non-Gaussian kernels, and it comes with guarantees on the filtering accuracy.

eess.IV↗

arXiv2017

Artifact reduction for separable non-local means

It was recently demonstrated [J. Electron. Imaging, 25(2), 2016] that one can perform fast non-local means (NLM) denoising of one-dimensional signals using a method called lifting. The cost of lifting is independent of the patch length, which dramatically reduces the run-time for large patches. Unfortunately, it is difficult to directly extend lifting for non-local means denoising of images. To bypass this, the authors proposed a separable approximation in which the image rows and columns are filtered using lifting. The overall algorithm is significantly faster than NLM, and the results are comparable in terms of PSNR. However, the separable processing often produces vertical and horizontal stripes in the image. This problem was previously addressed by using a bilateral filter-based post-smoothing, which was effective in removing some of the stripes. In this letter, we demonstrate that stripes can be mitigated in the first place simply by involving the neighboring rows (or columns) in the filtering. In other words, we use a two-dimensional search (similar to NLM), while still using one-dimensional patches (as in the previous proposal). The novelty is in the observation that one can use lifting for performing two-dimensional searches. The proposed approach produces artifact-free images, whose quality and PSNR are comparable to NLM, while being significantly faster.

cs.CV↗