Single-beam femtosecond laser pulses can surpass the diffraction limit of conventional focusing systems, enabling deep sub-waveguide single-spot modifications via nonlinear absorption at the focus. However, extending this capability to multi-beam parallel processing has been fundamentally limited by the diffraction limit (~λ/2). Herein, we experimentally and theoretically clarify that this limitation stems from spatial coherence and temporal sequence interference, inducing laser-matter interaction crosstalk, non-uniform modifications, and sparse duty-cycle structures. To address this, we propose a de-coherent parallel direct laser writing (Dc-PDLW) strategy, utilizing a patterned single pulse together with a de-coherent holographic algorithm (SSP-BM) to ensure multi-foci polarization orthogonality and eliminate spatial coherence. This method achieves single-shot fabrication of ultra-dense nanopore arrays, with 300 nm (~λ/4) resolution in crystals. We further demonstrate centimeter-scale 3D Pancharatnam-Berry phase plates and voluminous cipher sequences, realizing high-density 3D phase and polarization coding.

Light Field Salient Object Detection (LFSOD) aims to identify visually distinctive regions by leveraging the complementary spatial-angular information inherent in 4D light field imagery. A major challenge lies in modeling angular dependencies and maintaining spatial coherence under sparse supervision. In this paper, we propose a weakly supervised network that consists of three interdependent modules. First, the Light Field Division (LFD) module utilizes epipolar geometry to extract direction-aware boundary features, enhancing the encoding of angular disparities. Second, the Light Field Spatial Association (LFSA) module anchors cross-view feature alignment using central-view point annotations, thereby enforcing spatial consistency and mitigating redundant representations. Third, the Light Field Saliency Local Clustering (LFLC) module introduces a joint boundary-appearance modeling strategy that integrates adaptive clustering with error-aware regularization to refine structural predictions. Experiments on three benchmark datasets show that our method consistently outperforms mainstream weakly supervised approaches. It also achieves superior performance compared to several fully supervised methods.

Vision Transformers have demonstrated remarkable performance in image classification and structural modeling; however, fixed patch partitioning and static positional encoding often disrupt spatial continuity, thereby limiting their ability to represent rotated structures and irregular boundary regions. To address these limitations, we propose the Moore-curve Vision Transformer (MViT), a Vision Transformer (ViT) framework based on a recursive Moore curve. The proposed framework comprises three key components. First, a multi-order fractal mapping is employed to optimize patch reordering and enhance the spatial coherence of the token sequence. Second, a 7×7 dynamic partitioning template together with a boundary compensation algorithm jointly optimizes dense structural representation and resolution adaptability. Third, a period-aware positional encoding module integrates fractal periodic parameters with convolutional features to align positional embeddings with the fractal traversal pattern. This design significantly enhances the structural adaptability of the model to complex image layouts. Experimental results show that MViT improves classification accuracy over ViT-B/16 by 0.52% and 0.31% on the CIFAR-100 and ImageNet-21k datasets, respectively, while also achieving noticeable improvements in PSNR and SSIM. Ablation and rotational perturbation experiments further confirm its robustness to rotation and localized focus variations. Moreover, MViT exhibits strong structural compatibility, maintaining stable performance across different Transformer backbones and diverse visual tasks.

Vision Transformers have demonstrated remarkable performance in image classification and structural modeling; however, fixed patch partitioning and static positional encoding often disrupt spatial continuity, thereby limiting their ability to represent rotated structures and irregular boundary regions. To address these limitations, we propose the Moore-curve Vision Transformer (MViT), a Vision Transformer (ViT) framework based on a recursive Moore curve. The proposed framework comprises three key components. First, a multi-order fractal mapping is employed to optimize patch reordering and enhance the spatial coherence of the token sequence. Second, a 7×7 dynamic partitioning template together with a boundary compensation algorithm jointly optimizes dense structural representation and resolution adaptability. Third, a period-aware positional encoding module integrates fractal periodic parameters with convolutional features to align positional embeddings with the fractal traversal pattern. This design significantly enhances the structural adaptability of the model to complex image layouts. Experimental results show that MViT improves classification accuracy over ViT-B/16 by 0.52% and 0.31% on the CIFAR-100 and ImageNet-21k datasets, respectively, while also achieving noticeable improvements in PSNR and SSIM. Ablation and rotational perturbation experiments further confirm its robustness to rotation and localized focus variations. Moreover, MViT exhibits strong structural compatibility, maintaining stable performance across different Transformer backbones and diverse visual tasks.

Accurate cropland mapping in data-scarce regions remains challenging due to limited field data, strong interannual climatic variability, and heterogeneous cropping systems. This study proposes an NDVI-based training sample migration framework that transfers labeled samples from reference years in irrigated and rainfed agricultural systems to a target year using time-series similarity analysis. Ten similarity metrics representing geometric, temporal alignment, and correlation-based families were systematically evaluated to identify optimal thresholds and robust hybrid combinations for stable cropland transfer. The migrated samples were used to train a Random Forest classifier to generate binary cropland maps for 2021. Independent validation yielded overall accuracies of 86% in Kazakhstan and 95% in Uzbekistan. Comparisons with global cropland products (WorldCereal 2021 and WorldCover 2021) demonstrated improved spatial coherence and reduced misclassification, particularly in semi-arid environments. The proposed framework extends the temporal utility of existing labeled datasets and supports scalable cropland mapping without the need for repeated annual field surveys.

Ultrafast nonlinear ultrasound imaging of gas vesicles (GVs) promises high-sensitivity biomolecular visualization for applications such as targeted molecular imaging and real-time tracking of gene expression. However, separating GV-specific signals from tissue background remains challenging due to tissue clutter and limitations of current methods, which require complex transmit schemes and suffer from incomplete tissue suppression. This study aims to develop and validate harmonic amplitude-modulated singular value decomposition (HAM-SVD), a novel technique that represents a shift from current GV imaging methods by exploiting the unique nonlinear pressure-dependence of the GV harmonic signature.
Approach: HAM-SVD employs single-cycle plane waves transmitted at 9.6 MHz across five tilted angles at a pulse repetition frequency of 2500 Hz, under four duty cycles with alternating polarity. Beamformed data are reshaped into a space-pressure Casorati matrix and decomposed via singular value decomposition (SVD). Tissue background is suppressed by discarding the first (weakly nonlinear tissue) and lowest (noise) singular modes, yielding images comprised solely of pressure-dependent second-harmonic GV signals. The method was validated through numerical simulations, in vitro phantom experiments, and in vivo rat lower limb imaging.
Main results: HAM-SVD achieved a signal-to-background ratio (SBR) of 19.16 ± 1.63 dB in vivo, significantly outperforming pulse inversion (14.19 ± 1.41 dB) and AM-SVD (15.79 ± 1.38 dB). Simulation and phantom studies demonstrated superior spatial coherence in singular vector decomposition and reduced nonlinear artifacts compared to AM-SVD. HAM-SVD enables wide-field, ultrafast imaging without complex transmit sequences while maintaining robust tissue clutter suppression across varying pressure levels.
Significance: By combining harmonic imaging with AM-SVD's adaptive clutter filtering, HAM-SVD overcomes limitations of conventional nonlinear techniques, including depth restrictions and incomplete tissue cancellation. This approach enhances molecular imaging specificity for GVs and holds translational potential for ultrasound localization microscopy of slow-flowing contrast agents and preclinical disease-targeted molecular imaging.

Hyperspectral images present significant challenges for conventional dimensionality reduction methods due to their high dimensionality, spectral redundancy, and complex spatial–spatial dependencies. While kernel-based sparse representation methods have shown promise in handling spectral non-linearities, they often fail to preserve spatial consistency and semantic discriminability during feature transformation. To address these limitations, we propose a novel semantic-guided kernel low-rank sparse preserving projection (SKLSPP) framework. Unlike previous approaches that primarily focus on spectral information, our method introduces three key innovations: a semantic-aware kernel representation that maintains discriminability through label constraints, a spatially adaptive manifold regularization term that preserves local pixel affinities in the reduced subspace, and an efficient optimization framework that jointly learns sparse codes and projection matrices. Extensive experiments on benchmark datasets demonstrate that SKLSPP achieves superior performance compared to state-of-the-art methods, showing enhanced feature discrimination, reduced redundancy, and improved robustness to noise while maintaining spatial coherence in the dimensionality-reduced features.

Scaling up bedload monitoring: a passive acoustic approach for large river systems.

Spatially regularized super-resolved constrained spherical deconvolution (SR2-CSD) of diffusion MRI data.

The optical solitons have been of considerable interest for a long time because of the important applications, such as all-optical information processing (e.g. all-optical switch, and all-logic gates, etc.), optical manipulation and beam control, etc. It was shown that an annular optical soliton may be formed when a fully coherent vortex beam propagates in strongly nonlocal nonlinear media (SNNM). The annular optical soliton with vortex has more advantages in applications than the Gaussian-like optical soliton without vortex. In practice, partially coherent beams are often encountered, and the partial coherence is one of the main features of laser beams. However, when a partially coherent vortex beam propagates in SNNM, an optical soliton cannot be formed due to partial coherence. The aim of this paper is to find a kind of partially coherent vortex soliton.<br>Based on the extended diffraction integral principle together with the ABCD matrix of SNNM, the analytical propagation formula of twisted partially coherent vortex（TPCV）beams in SNNM is derived in this paper. It is found that an annular optical soliton may be formed in SNNM because of the twist feature of TPCV beams, even if the spatial coherence is extremely low. The conditions for the formation of annular optical solitons of TPCV beams in SNNM are also given in this paper. In addition, it is shown that the intensity and the gradient force of annular optical solitons increase as the partial coherence of TPCV beams decreases, which may be applied in optical manipulation.<br>On the other hand, under certain conditions, an optical soliton may also be formed, when a TPCV beam and a twisted Gaussian Schell-model (TGSM) beam are combined coaxially and incoherently in SNNM. The conditions for the formation of optical solitons of the combined beams in SNNM are independent of the beam coherence degree, the topological charge, and the proportion of sub-beam power. Furthermore, the gradient force can be manipulated by the beam coherence degree, and the profile of optical solitons can be manipulated by the topological charge and the proportion of sub-beam power. The results obtained in this paper is useful for optical manipulation, material processing, and beam control.

A liquid-nitrogen-cooled silicon channel-cut monochromator was developed and experimentally evaluated under high-thermal-load conditions. Under the maximum load of 417 W, the first reflecting surface exhibited a concave deformation, resulting in only an 11% reduction in vertical beam size at 16 m downstream. The deformation radius was estimated at 510 m. Despite the deformation, no significant changes were observed in the angular profile or intensity of the monochromatic beam. Interference fringes caused by edge diffraction at an upstream slit confirmed excellent preservation of spatial coherence. For the stability test of the monochromator, intensity fluctuation of the monochromatic beam was monitored and linearly fitted with upstream beam-position monitor signals, which were synchronously acquired. A high correlation (R2 = 0.95) confirmed that the inherent stability of the channel-cut design remained under cryogenic cooling. Additionally, a double channel-cut monochromator configuration for fixed-exit beam operation was tested and produced the expected output beam intensity. These results confirm the feasibility of using channel-cut monochromators as high-stability high-heat-load-tolerant optical elements for next-generation synchrotron beamlines.

Research on the Doppler ground simulation system based on spatial coherence laser communication link

This study addresses the challenge of effectively in- tegrating Convolutional Neural Networks (CNNs) and Trans- formers for complex vision tasks. While existing hybrids lever- age CNNs local feature extraction and Transformers global modeling capabilities, they typically employ simplistic feature transitions that fail to optimize the synergy between these archi- tectures. To bridge this gap, we propose the Taylor-Expansion- Enhanced CNN-based Encoder (TEECE) a novel framework that embeds trainable Taylor series expansion modules directly within the CNN backbone. These periodically integrated lay- ers mathematically recompose features across network stages, enhancing representational capacity while preserving spatial coherence. Unlike auxiliary Taylor approximations, our approach fundamentally rearchitects feature propagation to create a unified local-to-global representation hierarchy. Experimental results on public dataset show promising enhance- ments in accuracy and generalisation, our model achieving the best accuracy of 96.5% in image classification tasks. Highlighting the effectiveness of our proposed method and its potential for broader application in complex data processing tasks. By establishing a new paradigm for mathematically grounded vision architectures, TEECE offers both performance break- throughs and a reusable foundation for complex data processin

Historic religious cities are living examples of cultural landscapes where spiritual traditions, heritage, and visitor experiences combine to demonstrate a timeless experience. It is very challenging to achieve balance among the demands of mass pilgrimage, heritage preservation, and urbanization. Govardhan, India is a Hindu religious town with historical significance. Millions of pilgrims travel to Govardhan every year to perform parikrama and take a holy dip in kunds. The quality of the visitor experience, spatial coherence, and heritage conservation are all at risk due to increasing urbanization and tourism. The study intends to create a paradigm for the sustainable management of religious heritage towns by evaluating the factors involving visitor perception, historical significance, and spatial visibility, employing a combination of computational methods and cognitive assessments. The study employed space syntax tools (visibility graph analysis and isovist area analysis) to quantify spatial significance (SS) and identify patterns of openness, congestion, and visibility along the parikrama route of Govardhan. By examining pilgrims’ cognitive surveys for openness, orientation, congestion, and spiritual impression, a cognitive index (CI) and heritage importance scores (HIS) have been developed. The computed spatial significance (SS) has been correlated with cognitive index (CI) and heritage importance (HIS) scores to create an experiential priority index (EPI). The study employs a mixed-method approach that incorporates heritage significance scoring, cognitive surveys, and spatial analytics, including methods such as the isovist area analysis and visibility graph analysis. In order to assess how spatial arrangement and intangible perceptions together influence visitor experience, these statistics are further combined using a composite experiential priority index (EPI). The findings show a strong correlation between spiritual orientation, visual connectivity, and spatial openness; locations such as ‘punchari ka lota temple’ and ‘kusum sarovar’ are high-priority nodes. In accordance with United Nation Sustainable Development Goals (SDGs) (11, 9, 12, 4.7, and 8.9), this research proposes a heritage impact assessment (HIA) framework that provides workable solutions for ecological restoration, heritage-sensitive zoning, sustainable pilgrimage management, and enhanced tourism.

Floodwater depth estimation is essential for disaster response and infrastructure planning yet remains challenging in urban areas with limited gage and hydrological data. This study presents a deep learning-based framework grounded in the hydrostatic equilibrium principle to estimate flood depth using a remote sensing approach. A series of ResNet architectures were trained and evaluated under two different scenarios: (a) a baseline model input using LiDAR-derived DTM and flood extent, and (b) an enhanced model incorporating additional terrain features such as slope, curvature, and Topographic Wetness Index (TWI). The results demonstrate that ResNet18 outperformed deeper models, achieving an RMSE of 0.71 ft, Huber Loss of 0.28 ft, MAE of 0.23 ft, SSIM of approximately 99% and R-Squared of approximately 94% under the enhanced scenario. Inclusion of terrain predictors led to significant improvements in prediction accuracy and spatial coherence. They improved Huber Loss by 28%, RMSE by 13%, and MAE by 21%. However, when applied to an unseen peri-urban catchment, model performance declined (RMSE = 1.95 ft), mainly due to limited and temporally misaligned ground truth data, and differences in spatial characteristics. Despite these limitations, ResNet18 generalizes well, mapping flood depth in unseen catchments, and demonstrates the potential for rapid assessments in data-scarce regions.

A Poincaré-like sphere for spatial coherence characteristics leading to “visual algorithms” is introduced. Deterministic and random Jones transformations, as well as coherence optimization between lights at two spatial locations, can be apprehended with simple geometric transformations analogous to the ones used with the standard Poincaré sphere for polarization. The joint representation of polarization and coherence characteristics in a single global polarization intrinsic coherence Poincaré sphere allows one to easily identify remarkable physical situations.

Smart pixels: Interpretable active dictionary learning with spatial coherence regularization for hyperspectral image classification

We propose a technique for measuring the spatial coherency matrix of a wavefront based on a single-port intensity detector. Our method relies on multiplexing the incident wavefront across a series of known realizations of a reconfigurable complex system onto the intensity detector. We consider a multi-port chaotic cavity with partially reconfigurable boundary conditions as an embodiment of the reconfigurable complex system. This matches recent experiments with 3D chaotic cavities whose walls are partially covered with a programmable metasurface. We formulate a system model that rigorously accounts for multiple scattering based on multi-port network theory. Then, we numerically validate the principle. The appeal of our technique lies in the low hardware complexity.

Abstract We describe a monolithic interferometer for spatial coherence measurements of both classical and quantum light sources. The design enables measurements on both a PDC-based quantum source and a classical thermal source, using two identical calcite crystals to control beam alignment via birefringence. The monolithic structure ensures inherent stability. Spatial coherence is measured through temporal interferograms and spectral analysis, with both methods showing close agreement with theoretical predictions. The system is robust and performs reliably for both quantum and classical light. Its design enables automated, rapid coherence measurements across different source types.

Private educational institutions located within residential neighbourhoods are devices that often lead to urban transformation at a fast pace. Unfortunately, the majority of residents do not control these changes. Expansion around institutions like Minhaj University, University of Management and Technology (UMT) and other in Lahore has changed its land use pattern adversely. The surrounding areas has become congestedd and has started acting as a magnet for informal trade, which is gradually leading to decline in its identity as a residential zone. This research traces the changes and asks the question of how placemaking can be used to recover spatial coherence and inclusivity in the academic district after the gentrification process. The research used satellite imagery (2005, 2025), mental mapping, field observations, and stakeholder interviews to gauge the size of the area covered by the park, which it had increased seven times. Moreover, the research reported a simultaneity of commercial pressures. The research found that there are continuous conflicts between pedestrians and vehicles, while students have insufficient public spaces. Furthermore, vendor activities remain unregulated, and conversation of residential units into hostels has become widespread. The study offers a detailed account of gentrification caused by education in Pakistan and the strategies of design to democratize the redevelopment of similar high, pressure university areas.