arXiv subjects

Buddhi Wijenayake

Publications and source records attributed to Buddhi Wijenayake.

9 recordsLinked to original sources

arXiv2026

Structure over Depth: A Single-Block Spatio-Temporal Transformer for Multi-Entity Reasoning

Modeling multi-entity temporal data requires capturing dependencies across entities, time, and their interactions. Transformer-based approaches perform well but often rely on deep stacks of layers to learn these heterogeneous dependencies implicitly, increasing computational cost. We revisit this problem from a structural perspective and decompose multi-entity temporal dynamics into three interaction types: spatial interactions among entities, temporal interactions across time, and cross interactions coupling the two domains. We propose a structured spatio-temporal transformer block that explicitly models all three within a single stage. It uses parallel spatial and temporal self-attention, followed by bidirectional cross-attention, and combines the outputs through learnable gated fusion. By directly encoding these complementary views, the model reduces the need for deep stacking. We evaluate the approach on video-based group activity recognition, skeleton-based human interaction analysis, and wearable sensor-based activity recognition. Despite its simplicity, the single structured Transformer block matches or outperforms deeper architectures with only 1.76M parameters. The results suggest that depth in prior models partly compensates for implicit and entangled interaction modeling, whereas explicit factorization offers a more efficient and transparent alternative. More broadly, this work supports a structure-first design principle: expressive multi-entity temporal reasoning can emerge by exposing interaction structure rather than relying on depth.

cs.AI↗

arXiv2026

Preprocessing Algorithm Leveraging Geometric Modeling for Scale Correction in Hyperspectral Images for Improved Unmixing Performance

Spectral variability significantly impacts the accuracy and convergence of hyperspectral unmixing algorithms. Many methods address complex spectral variability; yet large-scale distortions to the scale of the observed pixel signatures due to topography, illumination, and shadowing remain a major challenge. These variations often degrade unmixing performance and complicate model fitting. Because of this, correcting these variations can offer significant advantages in real-world GIS applications. In this paper, we propose a novel preprocessing algorithm that corrects scale-induced spectral variability prior to unmixing. By estimating and correcting these distortions to the scale of the pixel signatures, the algorithm produces pixel signatures with minimal distortions in scale. Since these distortions in scale (which hinder the performance of many unmixing methods) are greatly minimized in the output provided by the proposed method, the abundance estimation of the unmixing algorithms is significantly improved. We present a rigorous mathematical framework to describe and correct for scale variability and provide extensive experimental validation of the proposed algorithm. Furthermore, the algorithm's impact is evaluated across a wide range of state-of-the-art unmixing methods on two synthetic and two real hyperspectral datasets. The proposed preprocessing step consistently improves the performance of these algorithms, achieving error reductions of around 50%, even for algorithms specifically designed to handle spectral variability. This demonstrates that scale correction acts as a complementary step, facilitating more accurate unmixing with existing methods. The algorithm's generality, consistent impact, and significant influence highlight its potential as a key component in practical hyperspectral unmixing pipelines. The implementation code will be made publicly available upon publication.

eess.IV↗

arXiv2026

Mitigating Long-Tail Bias via Prompt-Controlled Diffusion Augmentation

Long-tailed class imbalance remains a fundamental obstacle in semantic segmentation of high-resolution remote-sensing imagery, where dominant classes shape learned representations and rare classes are systematically under-segmented. This challenge becomes more acute in cross-domain settings such as LoveDA, which exhibits an explicit Urban/Rural split with substantial appearance differences and inconsistent class-frequency statistics across domains. We propose a prompt-controlled diffusion augmentation framework that generates paired label-image samples with explicit control over semantic composition and domain, enabling targeted enrichment of underrepresented classes rather than indiscriminate dataset expansion. A domain-aware, masked, ratio-conditioned discrete diffusion model first synthesizes layouts that satisfy class-ratio targets while preserving realistic spatial co-occurrence, and a ControlNet-guided diffusion model then renders photorealistic, domain-consistent images from these layouts. When mixed with real data, the resulting synthetic pairs improve multiple segmentation backbones, especially on minority classes and under domain shift, showing that better downstream segmentation comes from adding the right samples in the right proportions. Source codes, pretrained models, and synthetic datasets are available at \href{https://buddhi19.github.io/SyntheticGen}{\texttt{buddhi19.github.io/SyntheticGen}}.

cs.CV↗

arXiv2026

A Controlled Benchmark of Visual State-Space Backbones with Domain-Shift and Boundary Analysis for Remote-Sensing Segmentation

Visual state-space models (SSMs) are increasingly promoted as efficient alternatives to Vision Transformers, yet their practical advantages remain unclear under fair comparison because existing studies rarely isolate encoder effects from decoder and training choices. We present a strictly controlled benchmark of representative visual SSM families, including VMamba, MambaVision, and Spatial-Mamba, for remote-sensing semantic segmentation, in which only the encoder varies across experiments. Evaluated on LoveDA and ISPRS Potsdam under a unified 4-stage feature interface and a fixed lightweight decoder, the benchmark reveals three main findings, intra-family scaling yields only modest gains, cross-domain generalization is strongly asymmetric, and boundary delineation is the dominant failure mode under distribution shift. Although visual SSMs achieve favorable accuracy-efficiency trade-offs relative to the controlled CNN and Transformer baselines considered here, the results suggest that future improvements are more likely to come from robustness-oriented design and boundary-aware decoding than from encoder scaling alone. By isolating encoder behavior under a unified and reproducible protocol, this study establishes a practical reference benchmark for the design and evaluation of future Mamba-based segmentation backbones

eess.IV↗

arXiv2026

Mamba-FCS: Joint Spatio- Frequency Feature Fusion, Change-Guided Attention, and SeK Loss for Enhanced Semantic Change Detection in Remote Sensing

Semantic Change Detection (SCD) from remote sensing imagery requires models balancing extensive spatial context, computational efficiency, and sensitivity to class-imbalanced land-cover transitions. While Convolutional Neural Networks excel at local feature extraction but lack global context, Transformers provide global modeling at high computational costs. Recent Mamba architectures based on state-space models offer compelling solutions through linear complexity and efficient long-range modeling. In this study, we introduce Mamba-FCS, a SCD framework built upon Visual State Space Model backbone incorporating, a Joint Spatio-Frequency Fusion block incorporating log-amplitude frequency domain features to enhance edge clarity and suppress illumination artifacts, a Change-Guided Attention (CGA) module that explicitly links the naturally intertwined BCD and SCD tasks, and a Separated Kappa (SeK) loss tailored for class-imbalanced performance optimization. Extensive evaluation on SECOND and Landsat-SCD datasets shows that Mamba-FCS achieves state-of-the-art metrics, 88.62% Overall Accuracy, 65.78% F_scd, and 25.50% SeK on SECOND, 96.25% Overall Accuracy, 89.27% F_scd, and 60.26% SeK on Landsat-SCD. Ablation analyses confirm distinct contributions of each novel component, with qualitative assessments highlighting significant improvements in SCD. Our results underline the substantial potential of Mamba architectures, enhanced by proposed techniques, setting a new benchmark for effective and scalable semantic change detection in remote sensing applications. The complete source code, configuration files, and pre-trained models will be publicly available upon publication.

eess.IV↗

arXiv2025

PoPStat-COVID19: Leveraging Population Pyramids to Quantify Demographic Vulnerability to COVID-19

Understanding how population age structure shapes COVID-19 burden is crucial for pandemic preparedness, yet common summary measures such as median age ignore key distributional features like skewness, bimodality, and the proportional weight of high-risk cohorts. We extend the PoPStat framework, originally devised to link entire population pyramids with cause-specific mortality by applying it to COVID-19. Using 2019 United Nations World Population Prospects age-sex distributions together with cumulative cases and deaths per million recorded up to 5 May 2023 by Our World in Data, we calculate PoPDivergence (the Kullback-Leibler divergence from an optimised reference pyramid) for 180+ countries and derive PoPStat-COVID19 as the Pearson correlation between that divergence and log-transformed incidence or mortality. Optimisation selects Malta's old-skewed pyramid as the reference, yielding strong negative correlations for cases (r=-0.86, p<0.001, R^2=0.74) and deaths (r=-0.82, p<0.001, R^2=0.67). Sensitivity tests across twenty additional, similarly old-skewed references confirm that these associations are robust to reference choice. Benchmarking against eight standard indicators like gross domestic product per capita, Gini index, Human Development Index, life expectancy at birth, median age, population density, Socio-demographic Index, and Universal Health Coverage Index shows that PoPStat-COVID19 surpasses GDP per capita, median age, population density, and several other traditional measures, and outperforms every comparator for fatality burden. PoPStat-COVID19 therefore provides a concise, distribution-aware scalar for quantifying demographic vulnerability to COVID-19.

stat.AP↗

arXiv2025

Precision Spatio-Temporal Feature Fusion for Robust Remote Sensing Change Detection

Remote sensing change detection is vital for monitoring environmental and urban transformations but faces challenges like manual feature extraction and sensitivity to noise. Traditional methods and early deep learning models, such as convolutional neural networks (CNNs), struggle to capture long-range dependencies and global context essential for accurate change detection in complex scenes. While Transformer-based models mitigate these issues, their computational complexity limits their applicability in high-resolution remote sensing. Building upon ChangeMamba architecture, which leverages state space models for efficient global context modeling, this paper proposes precision fusion blocks to capture channel-wise temporal variations and per-pixel differences for fine-grained change detection. An enhanced decoder pipeline, incorporating lightweight channel reduction mechanisms, preserves local details with minimal computational cost. Additionally, an optimized loss function combining Cross Entropy, Dice and Lovasz objectives addresses class imbalance and boosts Intersection-over-Union (IoU). Evaluations on SYSU-CD, LEVIR-CD+, and WHU-CD datasets demonstrate superior precision, recall, F1 score, IoU, and overall accuracy compared to state-of-the-art methods, highlighting the approach's robustness for remote sensing change detection. For complete transparency, the codes and pretrained models are accessible at https://github.com/Buddhi19/MambaCD.git

eess.IV↗

arXiv2025

Enhanced SCanNet with CBAM and Dice Loss for Semantic Change Detection

Semantic Change Detection (SCD) in remote sensing imagery requires accurately identifying land-cover changes across multi-temporal image pairs. Despite substantial advancements, including the introduction of transformer-based architectures, current SCD models continue to struggle with challenges such as noisy inputs, subtle class boundaries, and significant class imbalance. In this study, we propose enhancing the Semantic Change Network (SCanNet) by integrating the Convolutional Block Attention Module (CBAM) and employing Dice loss during training. CBAM sequentially applies channel attention to highlight feature maps with the most meaningful content, followed by spatial attention to pinpoint critical regions within these maps. This sequential approach ensures precise suppression of irrelevant features and spatial noise, resulting in more accurate and robust detection performance compared to attention mechanisms that apply both processes simultaneously or independently. Dice loss, designed explicitly for handling class imbalance, further boosts sensitivity to minority change classes. Quantitative experiments conducted on the SECOND dataset demonstrate consistent improvements. Qualitative analysis confirms these improvements, showing clearer segmentation boundaries and more accurate recovery of small-change regions. These findings highlight the effectiveness of attention mechanisms and Dice loss in improving feature representation and addressing class imbalance in semantic change detection tasks.

eess.IV↗

arXiv2025

Devising PoPStat: A Metric Bridging Population Pyramids with Global Disease Mortality

Understanding the relationship between population dynamics and disease-specific mortality is central to evidence-based health policy. This study introduces two novel metrics, PoPDivergence and PoPStat, one to quantify the difference between population pyramids and the other to assess the strength and nature of their association with the mortality of a given disease. PoPDivergence, based on Kullback-Leibler divergence, measures deviations between a countrys population pyramid and a reference pyramid. PoPStat is the correlation between these deviations and the log form of disease-specific mortality rates. The reference population is selected by a brute-force optimization that maximizes this correlation. Utilizing mortality data from the Global Burden of Disease 2021 and population statistics from the United Nations, we applied these metrics to 371 diseases across 204 countries. Results reveal that PoPStat outperforms traditional indicators such as median age, GDP per capita, and Human Development Index in explaining the mortality of most diseases. Noncommunicable diseases (NCDs) like neurological disorders and cancers, communicable diseases (CDs) like neglected tropical diseases, and maternal and neonatal diseases were tightly bound to the underlying demographic attributes whereas NCDs like diabetes, CDs like respiratory infections and injuries including self-harm and interpersonal violence were weakly associated with population pyramid shapes. Notably, except for diabetes, the NCD mortality burden was shared by constrictive population pyramids, while mortality of communicable diseases, maternal and neonatal causes and injuries were largely borne by expansive pyramids. Therefore, PoPStat provides insights into demographic determinants of health and empirical support for models on epidemiological transition. Code and scripts: https://github.com/Buddhi19/DevisingPoPStat.git

stat.AP↗