As a judicious correspondence to the classical maxcut, the anti-Cheeger cut has more balanced structure, but few numerical results on it have been reported so far. In this paper, we propose a continuous iterative algorithm (CIA) for the anti-Cheeger cut problem through fully using an equivalent continuous formulation. It does not need rounding at all and has advantages that all subproblems have explicit analytic solutions, the objective function values are monotonically updated and the iteration points converge to a local optimum in finite steps via an appropriate subgradient selection. It can also be easily combined with the maxcut iterations for breaking out of local optima and improving the solution quality thanks to the similarity between the anti-Cheeger cut problem and the maxcut problem. The performance of CIAs is fully demonstrated through numerical experiments on G-set from two aspects: one is on the solution quality where we find that the approximate solutions obtained by CIAs are of comparable quality to those by the multiple search operator heuristic method; the other is on the computational cost where we show that CIAs always run faster than the often-used continuous iterative algorithm based on the rank-two relaxation.
Injection of bacteria with petroleum degrading capability into contaminated sites is one of the most cost-effective and environmental friendly strategies for the successful remediation of petroleum-contaminated groundwater. The successful in-situ bioremediation of petroleum contamination in subsurface is greatly impacted by the mobile/retention performance of petroleum-degrading bacteria in porous media, which yet is not well understood. The present study systematically investigated the mobile performance of petroleum-degrading strains in porous media with petroleum contamination under environmentally relevant solution and flow conditions. We found that although the mobile performance of petroleum-degrading bacteria was similar to petroleum non-degrading bacteria in uncontaminated porous media, bacteria containing different petroleum degrading function yet exhibited opposite transport behaviors in petroleum contaminated porous media. Enhanced mobility in porous media with petroleum contamination was achieved for petroleum non-degrading bacteria, while reduced mobility was obtained for petroleum-degrading bacteria. Combining the batch adsorption experiments, capillary chemotaxis assays, in-situ microfluidic chamber experiments together with theoretical calculation, we found that the opposite mobile performance observed for bacteria containing different petroleum degrading functions could be mainly attributed to their different chemotactic responses towards petroleum with negative and positive chemotaxis response respectively for non- and petroleum-degrading bacteria. Clearly, pollutant-degrading bacteria exhibited different mobile performance from non-degrading bacteria in contaminated porous media. The previous findings achieved from the model bacteria without pollutant-degrading capability could not be simply used to predict the mobile performance of pollutant-degrading bacteria. To ensure the successful implementation of in-situ bioremediation, the mobility of pollutant-degrading bacteria in contaminated porous media should be fully understood.
Mineral precipitation is ubiquitous in natural and engineered environments, such as carbon mineralization, contaminant remediation, and oil recovery in unconventional reservoirs. The precipitation process continuously alters the medium permeability, thereby influencing fluid transport and subsequent reaction kinetics. The diversity of preferential precipitation zones controls flow and transport efficiency as well as the capacity of mineral sequestration and immobilization. Taking barite precipitation as an example, previous studies have examined this process in porous and/or fractured media, but pore-scale mechanisms under varying flowing and geochemical conditions remain unexplored. In this study, we conducted real-rock microfluidic experiments to investigate the precipitation dynamics within a fractured porous system. Direct observations of the evolution of the porous structure and flow channel and quantifications of barite precipitation dynamics using X-ray diffraction (XRD) and scanning electron microscopy with energy-dispersive X-ray spectroscopy (SEM-EDS), revealed two distinct precipitation regimes: precipitation on the fracture surface (regime I) and precipitation in the alteration zone (regime II). Through theoretical analysis of the rate of advection and nucleation, we defined a dimensionless number Da above which regime I occurs and regime II prevails otherwise. At the large Da number, when the precipitation rate is large compared with the flow rate, precipitation on the fracture surface is favored. As the precipitation regimes are expected to impact differently the permeability of the fractured porous media, the mass transfer across matrix and fractures, and the spatial distributions of coprecipitated contaminants, our work sheds light on accurately modeling reactive transport in fractured porous media across diverse applications.
Impurity-containing iron hydroxides, abundant in many natural and engineered soil and aqueous environments, control the fate and transport of multiple aqueous contaminants. Fe(III) hydroxide was reported to simultaneously detoxicate As(III) and Cr(VI). However, the mechanisms and reaction intermediates are not clear, and the effects of impurities in ferrihydrite were far from being well understood. Here, Cr(III)-incorporated Fe(III) hydroxides were precipitated from acidic solutions (pH ∼ 3.0) with varied Fe(III)/Cr(III) molar ratios (10 : 0 to 8 : 2) for simultaneous removal of As(III) and Cr(VI). Multiple characterization techniques were combined to investigate the effects of Cr-incorporation on the size, band gap, adsorption, and catalytic efficiency of Fe hydroxides. With the amounts of Cr-incorporation increasing, the particle size of Fe hydroxides rapidly decreased (from 16.7 to 6.0 nm), and the removal of total As/Cr increased, as the Cr-incorporated Fe hydroxides with smaller size had larger surface area, promoting As/Cr removal by adsorption. Based on As/Cr speciation analysis of both aqueous and solid phases, the molar ratios of the oxidized As(III) (88%) to reduced Cr(VI) (∼56%) were calculated to be ∼1.5, indicating that the coupled redox conversion was the dominant removal mechanism over As(III)/Cr(VI) adsorption and As(III) oxidation. Intermediate characterization and molecular simulation found that Cr-incorporation promoted the early formation of H2O2 and Cr(V) intermediates, and enhanced the adsorption of reaction intermediates on Cr-incorporated Fe hydroxides, thus promoting their catalytic efficiency for coupled As(III)/Cr(VI) redox reactions.
The Transformer model, particularly its cross-attention module, is widely used for feature fusion in target sound extraction which extracts the signal of interest based on given clues. Despite its effectiveness, this approach suffers from low computational efficiency. Recent advancements in state space models, notably the latest work Mamba, have shown comparable performance to Transformer-based methods while significantly reducing computational complexity in various tasks. However, Mamba’s applicability in target sound extraction is limited due to its inability to capture dependencies between different sequences as the cross-attention does. In this paper, we propose CrossMamba for target sound extraction, which leverages the hidden attention mechanism of Mamba to compute dependencies between the given clues and the audio mixture. The calculation of Mamba can be divided to the query, key and value. We utilize the clue to generate the query and the audio mixture to derive the key and value, adhering to the principle of the cross-attention mechanism in Transformers. Experimental results from two representative target sound extraction methods validate the efficacy of the proposed CrossMamba
ABSTRACT An increasing body of research has investigated the role of enjoyment in second language acquisition (SLA); however, few studies have explored whether learners of Chinese as a second/foreign language (CS/FL) experience enjoyment in learning Hanzi (Chinese characters) and how enjoyment impacts Hanzi recognition performance. To address this gap, a Hanzi Learning Enjoyment Scale was developed and administered to 446 Arabic CS/FL learners, 144 of whom also completed a Hanzi recognition test. Two key findings emerged. First, the results of factor analysis revealed four factors underlying Hanzi learning enjoyment: Hanzi culture, personal attitudes, teacher support, and personal fulfillment. Second, enjoyment did not emerge as a significant predictor of Hanzi recognition performance. Notably, the variance in Hanzi recognition scores explained by enjoyment ranked among the top three explanatory variables, comparable to the predictive power of years spent learning Chinese. This study concludes with theoretical insights into the construct of foreign language enjoyment (FLE) across different languages and language components, as well as practical recommendations for enhancing Hanzi instruction.
Bijdragen irregularly organizes a book debate. This time we chose Rudolf Mrázek, Amir Sjarifoeddin: Politics and Truth in Indonesia, 1907–1948 (2024). We invited Henk Schulte Nordholt, KanKan Xie and Faizah Zakaria to share their critical insights from this book, to which Rudolf Mrázek responds.
Existing diffusion models for low-light image enhancement typically incrementally remove noise introduced during the forward diffusion process using a denoising loss, with the process being conditioned on input low-light images. While these models demonstrate remarkable abilities in generating realistic high-frequency details, they often struggle to restore fine details that are faithful to the input. To address this, we present a novel detail-preserving diffusion model for realistic and faithful low-light image enhancement. Our approach integrates a size-agnostic diffusion process with a reverse process reconstruction loss, significantly enhancing the fidelity of enhanced images to their low-light counterparts and enabling more accurate recovery of fine details. To ensure the preservation of region- and content-aware details, we employ an efficient noise estimation network with a simplified channel-spatial attention mechanism. Additionally, we propose a multiscale ensemble scheme to maintain detail fidelity across diverse illumination regions. Comprehensive experiments on eight benchmark datasets demonstrate that our method achieves state-of-the-art results compared to over twenty existing methods in terms of both perceptual quality (LPIPS) and distortion metrics (PSNR and SSIM). The code is available at: https://github.com/CSYanH/DePDiff.
