搜索结果：proposal

Circularly polarized organic afterglow (CPOA) materials have garnered considerable interest for their potential in information encryption, 3D displays, and sensing technologies. However, realizing CPOA materials that simultaneously offer high efficiency, long lifetime, high color purity, and large dissymmetry factor (glum) remains a significant challenge. Herein, an effective CP-hyperafterglow design strategy that rationally integrates narrowband hyperafterglow polymers with cholesteric liquid crystal matrices is proposed. The resulting polymeric films deliver multicolor narrowband CP-hyperafterglow emission with high photoluminescence quantum yields of up to 81%, emission bandwidths as narrow as ∼40 nm, ultralong lifetimes reaching 957 s, and maximum |glum| values of up to 1.3. Benefiting from these photophysical merits, various applications, including information encoding, multilevel encryption, and chiral display, are demonstrated. These findings offer a simple and reliable route toward high-performance CP-hyperafterglow, advancing the development of chiral optoelectronic materials and applications.

This paper argues that fungal mycelial networks exhibit minimal cognition through memory-integrated adaptive regulation. Drawing on cybernetic and enactivist frameworks, I develop a non-representational account of memory as the organism's capacity to modulate behavior based on temporally extended environmental coupling. I propose four operational criteria for minimal cognition: feedback-guided regulation of behavior, maintenance of internal viability conditions, structural modulation based on past environmental interactions, and plasticity across time scales that supports anticipatory adaptivity. Empirical evidence demonstrates that fungi meet all four criteria through distributed memory mechanisms: fungal networks exhibit directional regrowth toward previously encountered resources even after spatial displacement, stress priming persists across multiple cell divisions, Spitzenkörper-mediated directional persistence in constrained environments, and transgenerational memory through spore imprinting. These findings challenge representationalist assumptions in cognitive science by showing that memory and cognition can emerge from morphodynamic, biochemical, and electrophysiological processes without necessary neural substrates or symbolic representations. Fungal cognition demonstrates that the organizational principles underlying cognition-feedback-driven adaptation, norm-preservation, and historical coupling-can be realized in radically different material substrates, expanding our understanding of what counts as a cognitive system.

Computational protein design using machine learning models has advanced rapidly since the introduction of AlphaFold2. There is now a suite of tools that enable in silico design of proteins with desired structures and properties. Most design workflows require fitting a designed backbone with a sequence that stabilizes it, and many machine learning sequence design models have been proposed. These models are trained to recover the native sequence paired with a known structure, a task known as native sequence recovery (NSR). Here, we demonstrate the limitations of optimizing a sequence design model only for NSR. We show that NSR is often misaligned with more important metrics of model performance: the compatibility of the generated sequence with the desired fold and the ability of the model to predict the energetic effects of mutations. We introduce PottsMPNN, which is trained to generate a Potts energy function consisting of single-residue and residue-pair terms from a protein backbone, and we demonstrate that learning a Potts model reduces NSR but improves sequence generation and energy prediction. We also trained PottsMPNN with noised backbone structures and multiple sequence alignments. In tests on held-out data, NSR decreased, but the quality of the designed sequences and energy predictions improved. By demonstrating the limitations of optimizing for NSR and the effectiveness of strategies that avoid NSR overoptimization, our work advances sequence design and highlights future directions for the broader protein design field.

Current cross-domain few-shot semantic segmentation (CD-FSS) methods tend to overlook a fundamental yet domain-agnostic prior: the spatial correspondence between support and query images driven by the task itself. Unlike semantic similarity, this spatial correlation arises from the consistent structural layout of foreground objects across domains. To exploit this structural prior, we propose a novel frequency-spatial dual space adaptation (FDSA) framework, to learn domain-invariant structures and task-specific priors by jointly suppressing domain-specific redundancy in frequency domain and reinforcing geometric priors in spatial domain. Specifically, FDSA consists of two sequential modules, i.e., the frequency structural adapter (FSA) and the spatial geometry adapter (SGA). FSA performs image modulation in the frequency domain by emphasizing low-frequency foreground semantics and attenuating high-frequency noise, thus maintaining structural integrity of these input images. By contrast, SGA leverages handcrafted local descriptors to extract keypoints from both support and query images, generating Gaussian-based geometric priors that highlight desirable aligned regions. Additionally, we introduce spatial-guided SAM refinement (SSR) to extend our spatial geometric prior into the Segment Anything Model (SAM). SSR generates a soft Gaussian point prompt centered on the coarse mask, enabling SAM to refine segmentation masks without manual intervention. This integration effectively bridges task-specific localization with high-quality segmentation. Extensive experiments on four standard CD-FSS benchmarks demonstrate that our method achieves new state-of-the-art performance. Code is available at https://github.com/whales-zhang/FDSA.git.

In response to the situation where it is not allowed to stick CFRP cloth at the bottom of a concrete beam and stick it on both sides of the beam, this article analyzes the factors that affect the ultimate flexural bearing capacity of reinforced concrete beams reinforced with CFRP on the side, and provides a calculation method for the flexural bearing capacity of reinforced concrete beams reinforced with CFRP on the side; At the same time, for the convenience of calculation, this paper explores the comprehensive consideration of the tensile force of carbon fiber cloth pasted on the side and the corresponding correction factor ηf of the force arm, and analyzes it by fitting a quadratic trend function with the ratio of CFRP pasting height to beam height (hf/h). Based on this, the calculation methods for the bending capacity of carbon fiber cloth pasted on the bottom surface according to the "Code" and the bending capacity of carbon fiber cloth pasted on the bottom surface according to the quadratic trend function are proposed. Research has shown that using CFRP to reinforce reinforced concrete beams on the side can effectively improve the flexural bearing capacity. After comparative analysis, the calculation results of three calculation methods are in good agreement with the experimental values; The correction coefficient ηf increases with the increase of the ratio of the bonding height to the beam height (hf/h). When the ratio of the bonding height to the beam height (hf/h) exceeds 0.25, the value of the correction coefficient ηf increases significantly; Especially when the ratio of the pasting height to the beam height (hf/h) exceeds 0.5, it is recommended to calculate the flexural bearing capacity of carbon fiber cloth pasted on the bottom surface according to the proposed quadratic trend function for ηf; At the same time, it is recommended to consider the reduction of the cross-sectional area of carbon fiber cloth as compensation when determining the flexural bearing capacity of reinforced concrete beams with carbon fiber cloth pasted on the side according to the calculation of the beam bottom. In order to reduce errors, the utilization coefficient of ψf is no longer limited. Theoretical analysis shows that there are critical values for the bonding height and thickness of carbon fiber cloth used for reinforcement. When these exceed the critical value, the effect on enhancing load-bearing capacity becomes insignificant or even declines.

Diffusion models have emerged as a powerful framework for tasks like image controllable generation and dense prediction. However, existing models often struggle to capture underlying semantics (e.g., edges, textures, shapes) and effectively utilize in-context learning, limiting their contextual understanding and image generation quality. Furthermore, high computational costs and slow inference speeds hinder their real-time applications. To address these challenges, we propose Underlying Semantic Diffusion (US-Diffusion), an enhanced diffusion model that improves underlying semantics learning, computational efficiency, and in-context learning capabilities on multi-task scenarios. We introduce Separate & Gather Adapter (SGA), which decouples input conditions for different tasks while sharing the architecture, enabling better in-context learning and generalization across diverse visual domains. We also present a Feedback-Aided Learning (FAL) framework, which leverages feedback signals to guide the model in capturing semantic details and dynamically adapting to task-specific contextual cues. Furthermore, we propose a plug-and-play Efficient Sampling Strategy (ESS) for dense sampling at time steps with high-noise levels, which aims at optimizing training and inference efficiency while maintaining strong in-context learning performance. Experimental results demonstrate that US-Diffusion outperforms the state-of-the-art method, achieving an average reduction of 7.47 in FID on Map2Image tasks and an average reduction of 0.026 in RMSE on Image2Map tasks, while achieving approximately 9.45× faster inference speed. Our method also demonstrates superior training efficiency and in-context learning capabilities, excelling in new datasets and tasks, highlighting its robustness and adaptability across diverse visual domains. The source code will be released at https://github.com/dragon-cao/US-Diffusion.

Superconducting phases with exotic symmetries that differ from the underlying crystalline lattice are at the focus of superconductivity research. Yet, despite intense interest, detecting the order parameter symmetry and topology remains a major challenge. Real-space imaging near atomic impurities with scanning tunneling microscopy (STM) has been highly successful in revealing nodes of the superconducting gap, in particular in cuprate superconductors, however the order parameter phase winding has so far remained inaccessible by STM techniques. We demonstrate that STM can access this phase information by exploiting Young-type quasiparticle interference patterns generated by pairs of impurities acting as beam splitters. Superconducting order parameter tomography (SOPT), a technique proposed here, utilizes the response of real-space interference patterns of Bogoliubov quasiparticles to the controlled rotation of impurity configurations, allowing us to reconstruct the momentum space structure of the gap function [Formula: see text]. As a concrete example, we consider Strontium Ruthenate, whose superconducting order remains a subject of ongoing debate, and demonstrate how SOPT can distinguish between competing order parameter candidates. The Young's interference fringes, nodal directions, and rotating beams, detected by SOPT, encode information about both the nodes and phase winding of the superconducting order parameter. This method provides a broadly applicable route to identifying unconventional and topological superconductivity and establishes particle-hole interference as a new imaging modality for superconducting order.

Disorder-induced phenomena in quantum many-body systems pose a challenge for analytical and numerical approaches at relevant time and system scales. To reduce the cost of disorder sampling, we investigated quantum circuits initialized in states that form tunable superpositions over all disorder configurations, which in lattice gauge theories can be interpreted as superpositions over gauge sectors. On the experimentally accessible timescales, we observed localization in the absence of disorder in one and two dimensions: Perturbations failed to diffuse despite fully disorder-free evolution and initial states. However, entropy measurements revealed that superposition-prepared states fundamentally differ from those obtained by direct disorder sampling. Leveraging superposition, we propose an algorithm with a polynomial speedup in sampling disorder configurations, a long-standing challenge in many-body localization studies.

The development of gas sensors that combine high sensitivity and selectivity with stable operation at room temperature under high humidity remains a pivotal challenge for reliable breath analysis. This work presents a rational design and facile synthesis of three-dimensional hierarchical beaded Er3+-doped CeO2/ZnO (Er:CeO2-ZnO) heterojunction nanofibers (NFs) via a ZIF-8-templated electrospinning strategy. This integrated approach converges multiple material design principles. It employs a ZIF-8-derived porous ZnO scaffold for enhanced gas accessibility, in situ formed n-n heterojunctions for efficient charge separation, and Er3+-induced oxygen vacancies in CeO2 for optimized surface chemistry. Comprehensive characterization confirms the successful integration of these components and the creation of abundant active sites. The optimized sensor exhibits exceptional room-temperature triethylamine (TEA) sensing performance, including a high response, a low detection limit (56 ppb), fast response/recovery kinetics (16 s/58 s to 10 ppm TEA), and excellent selectivity. Remarkably, the sensor demonstrates outstanding long-term stability (over 120 days) and humidity-independent performance across a wide relative humidity range (25-90% RH). Kinetic analysis of the sensing transients reveals biphasic behavior, providing direct evidence for a dual-active-site synergistic mechanism involving oxygen vacancy-rich regions for gas adsorption/activation and heterojunction interfaces for charge separation. Furthermore, a dynamic self-refreshing mechanism is proposed to elucidate the exceptional humidity tolerance. This work provides fundamental insights into the design of multifunctional sensing materials and presents a highly promising candidate for reliable gas detection in complex, humid environments such as exhaled breath analysis.

Lung cancer lesion segmentation in two-dimensional computed tomography (2D CT) images remains challenging due to blurred boundaries, heterogeneous morphologies, and annotation uncertainty, leading to unreliable delineations and reduced clinical usability. To address this research gap, we propose a novel 2D CT lung cancer semantic segmentation framework, OncoSeg2D, which explicitly tackles boundary ambiguity and morphological distortion through two complementary modules. Specifically, an Uncertainty-aware Boundary Modeling (UBM) module probabilistically represents tumor edges via learnable mean-variance estimation and gradient-weighted sampling, while a Morphology-Preserving Regularization (MPR) module constrains the segmentation with curvature, compactness, and convexity priors to maintain global shape consistency. The framework integrates these designs with multi-scale feature extraction from a SegFormer backbone and requires no additional annotations or three-dimensional (3D) reconstruction. Experiments conducted on the Medical Segmentation Decathlon Challenge dataset and the lung cancer segmentation dataset demonstrate that OncoSeg2D achieves IoU scores of 0.865 and 0.788, mIoU scores of 0.881 and 0.799, and Dice Similarity Coefficients (DSC) of 0.923 and 0.816, consistently outperforming conventional CNN-based models and mainstream Transformer-based methods. Compared with the SegFormer baseline, the proposed method improves mIoU by 3.8% and 4.0% on the two datasets, respectively, while reducing the Hausdorff distance from 5.61 to 3.41 and from 7.26 to 5.28, indicating superior boundary refinement and stronger global shape consistency. These results verify that explicitly integrating uncertainty modeling and morphological priors yields both higher accuracy and better interpretability. Overall, the proposed framework not only enhances segmentation accuracy but also improves clinical interpretability and reliability, offering a promising solution for lung cancer diagnosis assistance and therapeutic outcome monitoring.

This article presents a new design of prescribed-time impulsive controllers for the prescribed-time stability (PTS) of high-order integrator systems with a single control input. To deal with the singularity of the control input matrix, a stability result is established based on the algebraic Riccati equation, which derives the explicit parametric control gains of impulsive control. Then, the prescribed-time impulsive control with such parametric control gains is designed to ascertain the convergence of the state transition matrix to a nilpotent matrix and achieve the PTS of high-order integrator systems by adjusting the parameters of control gains at impulsive instants. Aprescribed-time impulsive observer is also proposed for the PTS of the system with an accessible sampled single output. Finally, the validity of the theoretical results is verified by the trajectory tracking of the omnidirectional mobile robots via velocity regulation.

Metastatic hormone-sensitive prostate cancer (mHSPC) exhibits heterogeneous progression patterns, with early progression to metastatic castration-resistant prostate cancer (mCRPC) within 12 months indicating aggressive tumor biology and poor prognosis. Current risk stratification tools (CHAARTED, LATITUDE) offer limited individualized prediction. Machine learning approaches are increasingly applied to predict prostate cancer progression, but most models show modest performance (AUC 0.68-0.72), limited external validation, or require genomic variables unavailable in routine practice. This study aimed to develop and externally validate a novel RINH algorithm for predicting early mCRPC progression (≤ 12 months) using exclusively clinical variables, positioning it as a superior alternative to conventional ML classifiers. This multicenter study enrolled 412 patients with de novo mHSPC from seven Spanish academic centers using mixed retrospective-prospective data collection. Twenty clinical variables were recorded, including demographics, PSA, ISUP grade, metastatic localization, CHAARTED/LATITUDE classifications, and treatment modalities. Following RINH-based outlier exclusion (55 patients), 357 patients (29 with early progression, 8.1%) were used to train six ML algorithms: RINH, Logistic Regression, Linear Discriminant, Support Vector Machine, Random Forest, and Subspace Discriminant. A two-tiered validation strategy integrated stratified fivefold cross-validation across all centers and formal external validation using center 1 (n = 121, 19 events) for training and centers 2-7 (n = 207, 10 events) for independent testing. Performance metrics included AUC, sensitivity, specificity, accuracy, and F1-score. Artificial intelligence and machine learning (ML) are transforming oncology, promising personalized risk stratification beyond traditional clinical criteria. In metastatic hormone-sensitive prostate cancer (mHSPC), early progression to castration resistance (mCRPC) within 12 months signals aggressive biology and poor prognosis, yet current tools (CHAARTED, LATITUDE) offer limited individualized prediction. Multiple ML models have been proposed with variable success: most achieve modest performance (AUC 0.68-0.72), lack robust external validation, or rely on genomic variables inaccessible in routine practice. We propose a novel approach using the Rivality Index Neighborhood (RINH) algorithm, demonstrating superior predictive capacity in an initial multicenter validation with exclusively clinical variables. This study provides rigorous multicenter external validation, advancing toward implementable precision oncology tools. The RINH algorithm achieves superior predictive performance for early mCRPC progression using exclusively clinical variables, representing a significant advance toward implementable risk stratification. However, low reliability scores in external validation underscore that excellent performance metrics alone do not guarantee stability. Before clinical deployment, validation in substantially larger cohorts with higher progression events is essential. If validated, this model could enable personalized, risk-adapted therapeutic strategies, refining patient selection for treatment intensification or de-escalation.

The study focuses on the design, development, optimization, and performance of novel porous composite media prepared using bulk utilization of Fly Ash (FA), Waste Tyre Rubber (WTR), Rice Husk Ash (RHA), and a minute amount of Cement (C) blended with water to eliminate the need for chemical additives. Various compositions were prepared and evaluated based on engineering properties, resulting in three optimized mixes: 80%FA + 20%C, 50%FA + 30%WTR + 20%C, and 50%FA + 15%WTR + 15%RHA + 20%C on weight basis. Morphological and physicochemical characterization was carried out using FESEM, EDX, and BET analyses. The optimized media were evaluated in a laboratory-scale wastewater treatment system consisting of an upflow anaerobic baffled reactor (UABR) followed by a baffled constructed wetland (CW), considering without and with wetland plants for six months at 24h HRT. The UABR unit achieved high removal efficiencies for suspended solids and moderates toward organic matter. Further CW with composite media improve the treatment performance mainly due to combined physical filtration, adsorption, microbial degradation, and plant-assisted nutrient uptake mechanisms. CW with developed media and wetland plants performed much better than conventional media with plants. TCLP analysis confirmed environmental safety, demonstrating the developed media as a sustainable and cost-effective alternative for treatment in a constructed wetland. This study proposes the design and development of a novel composite media for domestic wastewater treatment, using a unique combination of industrial and agricultural wastes (FA, WTR, and RHA) with minimal cement and no chemical additives. The research further validates the treatment efficiency, long-term stability, and environmental safety of the media when used in a constructed wetland system.

Graph neural networks (GNNs) have shown promising performance in solving both Boolean satisfiability (SAT) and maximum satisfiability (MaxSAT) problems due to their ability to efficiently model and capture the structural dependencies between literals and clauses. However, GNN methods for solving weighted MaxSAT problems remain underdeveloped. The challenges arise from the nonlinear dependency and sensitive objective function, which are caused by the nonuniform distribution of weights across clauses. In this article, we present HyperSAT, a novel neural approach that employs an unsupervised hypergraph neural network (HNN) model to solve weighted MaxSAT problems. Specifically, we propose a hypergraph representation for weighted MaxSAT instances to encode higher-order relationships between literals and clauses. A cross-attention mechanism and a shared representation constraint loss function are designed to capture the logical interactions between positive and negative literal nodes in the hypergraph, which effectively address the challenges posed by the uneven weight distribution. Compared with GNN-based SAT solvers that perform message passing only along pairwise connections, the proposed HNN enables multi-literal message passing within hyperedges, providing a more expressive mechanism to capture the weighted group interactions in complex clauses. Extensive experiments on various weighted MaxSAT datasets demonstrate that HyperSAT achieves better performance than state-of-the-art learning-based approaches, with average relative improvements ranging from 1.80% to 13.37%.

To improve interprofessional collaboration among occupational health professionals (OHPs) during the return-to-work (RTW) process, the Work Ability & Reintegration Description (WARD) instrument and a multidisciplinary guideline were developed based on the International Classification of Functioning, Disability and Health (ICF). The objectives of this study were [1]: to explore OHPs' experiences of challenges in interprofessional collaboration during sick leave and RTW guidance; and [2] to refine the WARD instrument to improve collaboration. Five consecutive focus groups were conducted with 11 OHPs, using the Design Thinking method as a user-centred approach to explore challenges and solutions related to interprofessional collaboration during sick leave and the RTW process. The Sunnybrook framework for interprofessional team collaboration was used as an analytical framework. Challenges related to interprofessional collaboration were identified in four of the six domains of the Sunnybrook framework: role delineation, communication, interprofessional values and ethics, and shared decision-making. OHPs highlighted that current instruments primarily focus on assessing limitations rather than possibilities for RTW and emphasized the importance of incorporating environmental and personal factors into the WARD instrument. This study identified key challenges in interprofessional collaboration during sick leave and the RTW process and proposes solutions through refinements to the WARD instrument and multidisciplinary guideline. By shifting the focus of work ability assessments from functional limitations toward RTW possibilities, integrating employees' and employers' perspectives, and fostering shared RTW goals, these refinements have the potential to enhance collaboration and improve RTW outcomes. Effective return-to-work guidance requires interprofessional collaboration based on a shared understanding of work capacity, incorporating medical, personal, and environmental factors.Using an ICF-based instrument that focuses on work possibilities rather than limitations can support more constructive dialogue and shared decision-making among rehabilitation stakeholders.Actively involving occupational health professionals, employees, and employers in return-to-work guidance promotes alignment of perspectives and strengthens engagement in the rehabilitation process.Design Thinking offers a practical, participatory approach for developing and refining rehabilitation tools that address real-world challenges in interprofessional collaboration.

Cooling towers are an important part of the thermal system in industries, where they are used to remove unwanted heat and help maintain the proper performance of the machines. Four machine learning algorithms, namely random forest, support vector machine (SVM), decision tree, and AdaBoost are proposed in this paper for the performance forecasting of cooling towers. for the performance forecasting of cooling towers. These models were built in Python with the help of the following operational parameters: inlet water temperature (32-41°C), ambient air temperature (14-32°C), and relative humidity (35-92%). All the essential performance measures like outlet water temperature, water losses, the effectiveness, and the second law efficiency were predicted and assessed by statistical indicators such as coefficient of determination (R²), root mean square error (RMSE) and mean absolute percentage error (MAPE). The SVM algorithm had the best predictive accuracy and lowest prediction errors of all the tested models with a value of R2 of 0.985 and RMSE of 1.25 kg/s. Parametric analysis had indicated that the increase in relative humidity between 35% and 92% decreased the evaporation losses by about 55-70% and makeup water demand by about 58-68%. Thermodynamic analysis further revealed that the second-law efficiency improved by approximately 65-75% as the ambient temperature increased. The results indicate that predictive modeling with machine learning offers a useful method in the optimization of cooling tower operation and minimizing water use in industrial systems.

This article presents a plug-and-play (PnP) distributed control framework for DC microgrids (DCmGs) to address scalability and reconfiguration issues under dynamic network topologies. First, a decentralized control architecture is developed based on the small-gain theorem, enabling PnP operations of distributed generation units (DGUs). Specifically, the plug-in operation only requires local and neighboring state feedback without relying on global information exchange, while the plug-out operation can be performed without interunit communication. In this way, controller deployment is decoupled from the global network topology, thereby reducing reconfiguration overhead and improving the modular scalability and operational flexibility of DCmGs. Second, an optimization-based controller synthesis method is proposed to minimize the coupling effects among DGUs. The proposed method facilitates distributed controller design for heterogeneous DGUs without requiring structured Lyapunov functions or free-weighting matrices, thereby simplifying the synthesis procedure. Theoretical analysis establishes the asymptotic stability and prescribed performance of the closed-loop system under PnP operations. Finally, simulation studies on a six-DGU DCmG prototype are provided to validate the effectiveness of the proposed framework.

Chemical exchange saturation transfer (CEST) is a powerful magnetic resonance imaging (MRI) technique for noninvasively probing both endogenous metabolites and exogenous contrast agents, with broad applications across various diseases. Although three-dimensional (3D) CEST MRI enables volumetric characterization of tissue heterogeneity, its clinical utility is hampered by long scan time because large volumetric datasets must be repeatedly acquired at multiple saturation frequency offsets. This limitation becomes more pronounced at high spatial resolutions. To address this challenge, we propose Wave-Co-CAIPI, a novel method that integrates Wave-CAIPI encoding with center-out reordering and keyhole sampling strategy, to substantially accelerate data acquisition for high-resolution 3D CEST MRI. Experiments conducted on in-vitro glutamate phantom and in-vivo human subjects at 5 Tesla demonstrate the feasibility and effectiveness of the method, achieving up to nine-fold k-space acceleration for 1 mm isotropic resolution in both amide proton transfer-weighted and glutamate-weighted 3D CEST imaging. These results highlight the potential of the proposed Wave-Co-CAIPI to enable highly accelerated, high-resolution 3D CEST MRI in clinical practice.

Referring Expression Segmentation (RES) requires models not only to locate objects specified by referring expressions accurately but also to predict complete and precise masks. Existing methods primarily focus on complex multimodal alignment for object grounding, often neglecting mask quality, which results in incomplete foreground regions and imprecise boundaries. To address these challenges, we propose DiffRES, a mask-generating framework based on Stable Diffusion (SD), designed to tackle the RES problem with a focus on achieving high-quality masks. DiffRES effectively mitigates the information leakage issue prevalent in existing generative dense prediction diffusion models, which allows the model to infer the target's position directly from noisy masks during training without understanding the text condition, leading to severe overfitting. Specifically, DiffRES directly guides SD with visual and linguistic information to generate target binary masks, fundamentally bypassing the information leakage issue. This approach enables efficient knowledge transfer from SD to the RES task, resulting in precisely localized binary masks with sharp and precise boundaries. Extensive experiments show that DiffRES surpasses current state-of-the-art traditional methods on boundary precision (APb) which is sensitive to mask quality, while also significantly outperforming all existing SD-based RES models across all metrics. Our code is publicly available at https://github.com/charon517-517/DiffRES.