Diffusion Model Research Articles

Network architecture for single image super‐resolution: A comprehensive review and comparison

AbstractSingle image super‐resolution (SISR) is a promising research direction in computer vision and image processing for improving the visual perception of low‐quality images. In recent years, deep learning algorithms have driven tremendous development in SR, and SR methods based on various network architectures have significantly improved the quality of reconstructed images. Although there has been a large amount of reviews focusing on SISR, few studies have focused specifically on network architectures for SISR. This paper aims to provide a systematic overview of the design ideas of SISR using multiple architectures, including Convolutional Neural Networks (CNN), Generative Adversarial Networks (GAN), Transformer, and Diffusion model. In addition, an experimental analysis and comparison of state‐of‐the‐art SR algorithms have been performed on publicly available quantitative and qualitative datasets. Finally, some future directions are discussed that may help other community researchers.

Quantum circuit synthesis with diffusion models

Quantum circuit synthesis with diffusion models

Leakage Source Location of Hazardous Chemicals Based on the Improved Gray Wolf Optimization Algorithm

To accurately determine the leakage source location and strength during gas leakage accidents, this study compares the concentration obtained from the diffusion model with that measured by the sensor and proposes an improved gray wolf optimization algorithm for leakage source location. This algorithm introduces two improvement strategies. First, a nonlinear convergence factor is introduced to balance the global and local searches of the algorithm. Second, a reverse learning operation is performed on the three individuals with the worst fitness in the contemporary population. The results showed that the location results based on the improved gray wolf optimization algorithm exhibited high accuracy and stability, could quickly and accurately locate the leakage source, and provided data support for emergency disposal of accidents.

The Comparison of the Effectiveness and Efficiency of Fine-Tuning Models on Stable Diffusion in Creating Concept Art

This research aims to overcome the limitations of the Stable Diffusion model in creating conceptual works of art, focusing on problem identification, research objectives, methodology and research results. Even though Stable Diffusion has been recognized as the best model, especially in the context of creating conceptual artwork, there is still a need to simplify the process of creating concept art and find the most suitable generative model. This research used three methods: Latent Diffusion Model, Dreambooth: fine-tuning Model, and Stable Diffusion. The research results show that the Dreambooth model produces a more real and realistic painting style, while Textual Inversion tends towards a fantasy and cartoonist style. Although the effectiveness of both is relatively high, with minimal differences, the Dreambooth model is proven to be more effective based on the consistency of FID, PSNR, and visual perception scores. The Dreambooth model is more efficient in training time, even though it requires more memory, while the inference time for both is relatively similar. This research makes a significant contribution to the development of artificial intelligence in the creative industries, opens up opportunities to improve the use of generative models in creating conceptual works of art, and can potentially drive positive change in the use of artificial intelligence in the creative industries more broadly.

Competition between physical search and a weak-to-strong transition rate-limits kinesin binding times.

The self-organization of cells relies on the profound complexity of protein-protein interactions. Challenges in directly observing these events have hindered progress toward understanding their diverse behaviors. One notable example is the interaction between molecular motors and cytoskeletal systems that combine to perform a variety of cellular functions. In this work, we leverage theory and experiments to identify and quantify the rate-limiting mechanism of the initial association between a cargo-bound kinesin motor and a microtubule track. Recent advances in optical tweezers provide binding times for several lengths of kinesin motors trapped at varying distances from a microtubule, empowering the investigation of competing models. We first explore a diffusion-limited model of binding. Through Brownian dynamics simulations and simulation-based inference, we find this simple diffusion model fails to explain the experimental binding times, but an extended model that accounts for the ADP state of the molecular motor agrees closely with the data, even under the scrutiny of penalizing for additional model complexity. We provide quantification of both kinetic rates and biophysical parameters underlying the proposed binding process. Our model suggests that a typical binding event is limited by ADP state rather than physical search. Lastly, we predict how these association rates can be modulated in distinct ways through variation of environmental concentrations and physical properties.

Emotion expression salience and racially biased weapon identification: A diffusion modeling approach.

Racial stereotypes are commonly activated by informational cues that are detectable in people's faces. Here, we used a sequential priming task to examine whether and how the salience of emotion (angry/scowling vs. happy/smiling expressions) or apparent race (Black vs. White) information in male face primes shapes racially biased weapon identification (gun vs. tool) decisions. In two experiments (Ntotal = 546) using two different manipulations of facial information salience, racial bias in weapon identification was weaker when the salience of emotion expression versus race was heightened. Using diffusion decision modeling, we tested competing accounts of the cognitive mechanism by which the salience of facial information moderates this behavioral effect. Consistent support emerged for an initial bias account, whereby the decision process began closer to the "gun" response upon seeing faces of Black versus White men, and this racially biased shift in the starting position was weaker when emotion versus race information was salient. We discuss these results vis-à-vis prior empirical and theoretical work on how facial information salience moderates racial bias in decision-making.

C-ShipGen: a conditional guided diffusion model for parametric ship hull design

Ship design is a complex design process that may take a team of naval architects many years to complete. Improving the ship design process can lead to significant cost savings, while still delivering high-quality designs to customers. A new technology for ship hull design is diffusion models, a type of generative artificial intelligence. Prior work with diffusion models for ship hull design created high-quality ship hulls with reduced drag and larger displaced volumes. However, the work could not generate hulls that meet specific design constraints. This paper proposes a conditional diffusion model that generates hull designs given specific constraints, such as the desired principal dimensions of the hull. In addition, this diffusion model leverages the gradients from a total resistance regression model to create low-resistance designs. Five design test cases compared the diffusion model to a design optimization algorithm to create hull designs with low resistance. In all five test cases, the diffusion model was shown to create diverse designs with a total resistance less than the optimized hull, having resistance reductions over 25%. The diffusion model also generated these designs without retraining. This work can significantly reduce the design cycle time of ships by creating high-quality hulls that meet user requirements with a data-driven approach.

Study on size effect and model optimization of methane desorption-diffusion in high rank coal

ABSTRACT Methane diffusion models are numerous, but their application, pros, and cons need to be summarized urgently, the link between particle size and methane desorption-diffusion is uncertain. Therefore, this paper carries out the methane desorption-diffusion experiments under multi-scale pressure and particle size conditions to study the effect of coal sample particle size on methane desorption-diffusion. Based on MATLAB, different diffusion models were fitted and comparatively analyzed, and the models were improved in view of their deficiencies. The results showed that the cumulative methane desorption amount and rate are inversely related to the particle size of coal. Comparative analysis of the models revealed that diffusion model under the third type of boundary conditions has a low fitting accuracy for full process of methane desorption, whereas it has a high fitting accuracy of 0.9831 for the first 1000 s of desorption data. Unipore diffusion model, bidisperse diffusion model and new model of dynamic diffusion coefficient have high fitting accuracy for the desorption data of high homogeneous coal, reaching above 0.97, but unipore diffusion model has only 0.5434 fitting accuracy for non-homogeneous coal samples. The fitting accuracy of improved classical unipore diffusion model for both homogeneous and non-homogeneous coals reaches 0.99, which is more extensive than that of unipore diffusion model, only take the first term can satisfy the accuracy requirement, so it is more convenient in the practical engineering application, and has certain reference significance for engineering calculations of gas content and prediction of gas protrusion.

The bass diffusion model: agent-based implementation on arbitrary networks

ABSTRACT The goal of this study is to model Bass diffusion and its extensions on complex networks, including scale-free networks with arbitrary power-law exponent and assortative degree correlations. For this purpose we employ a combination of the free software packages networkX and NetLogo. Some new results obtained in the agent-based simulations (and differing from those in mean-field approximation) are the following. The introduction of assortative correlations in scale-free networks has the effect of delaying the adoption peak in the Bass model, compared to the uncorrelated case. The peak time depends strongly also on the maximum degree effectively present in the network. For diffusion models with threshold on signed network, a high level of clustering tends to cause adoption blockades. By analysing statistical ensembles of assortative networks generated via Newman rewiring one observes a remarkable strong correlation between the average degree of first neighbours k ˉ nn ( k ) and the average clustering coefficient C ˉ ( k ) .

Enhancing therapeutic efficacy: sustained delivery of 5-fluorouracil (5-FU) via thiolated chitosan nanoparticles targeting CD44 in triple-negative breast cancer

Our current study reports the successful synthesis of thiolated chitosan-based nanoparticles for targeted drug delivery of 5-Fluorouracil. This process was achieved through the ionic gelation technique, aiming to improve the efficacy of the chemotherapeutic moiety by modifying the surface of the nanoparticles (NPs) with a ligand. We coated these NPs with hyaluronic acid (HA) to actively target the CD44 receptor, which is frequently overexpressed in various solid malignancies, including breast cancer. XRD, FTIR, SEM, and TEM were used for the physicochemical analysis of the NPs. These 5-Fluorouracil (5-FU) loaded NPs were evaluated on MDA-MB-231 (a triple-negative breast cell line) and MCF-10A (normal epithelial breast cells) to determine their in vitro efficacy. The developed 5-FU-loaded NPs exhibited a particle size within a favorable range (< 300 nm). The positive zeta potential of these nanoparticles facilitated their uptake by negatively charged cancer cells. Moreover, they demonstrated robust stability and achieved high encapsulation efficiency. These nanoparticles exhibited significant cytotoxicity compared to the crude drug (p < 0.05) and displayed a promising release pattern consistent with the basic diffusion model. These traits improve the pharmacokinetic profile, efficacy, and ability to precisely target these nanoparticles, offering a potentially successful anticancer treatment for breast cancer. However, additional in vivo assessments of these formulations are obligatory to confirm these findings.

IMNE: Maximizing influence through deep learning-based node embedding in social network

Influence Maximization (IM) is a critical problem in social network analysis and marketing. It involves identifying a subset of nodes in a social network whose activation or influence can lead to the maximal spread of information, ideas, or behaviors within the network. Although many approaches have been developed in the literature to deal with this problem, most of these approaches are ineffective in dealing with large-scale social networks due to free parameters and computational complexity. Embeddings are used to learn low-dimensional representations of nodes in a social network. These embeddings capture the structural and semantic information of nodes and their relationships within the network. By training deep learning models on graph-structured data, node embeddings can capture complex patterns and dependencies in social networks, enabling more effective downstream tasks such as IM. Accordingly, this paper proposes an efficient algorithm to address the IM problem in social networks using deep learning-based Node Embedding (IMNE), which includes shell decomposition, graph/node embedding, and search space reduction as well as the use of local structural features. Our approach combines the power of deep learning for representation learning with the rich structural information present in social networks to address the challenge of IM in complex and dynamic social networks. IMNE uses the Independent Cascade (IC) information diffusion model to determine the labels needed to train the model by calculating the influence of nodes. Experimental results on several real-world networks considering different performance metrics show that IMNE performs better compared to existing baseline and state-of-the-art methods.

Conditional Diffusion Model for Urban Morphology Prediction

Predicting urban morphology based on local attributes is an important issue in urban science research. The deep generative models represented by generative adversarial network (GAN) models have achieved impressive results in this area. However, in such methods, the urban morphology is assumed to follow a specific probability distribution and be able to directly approximate the distribution via GAN models, which is not a realistic strategy. As demonstrated by the score-based model, a better strategy is to learn the gradient of the probability distribution and implicitly approximate the distribution. Therefore, in this paper, an urban morphology prediction method based on the conditional diffusion model is proposed. Implementing this approach results in the decomposition of the attribute-based urban morphology prediction task into two subproblems: estimating the gradient of the conditional distribution, and gradient-based sampling. During the training stage, the gradient of the conditional distribution is approximated by using a conditional diffusion model to predict the noise added to the original urban morphology. In the generation stage, the corresponding conditional distribution is parameterized based on the noise predicted by the conditional diffusion model, and the final prediction result is generated through iterative sampling. The experimental results showed that compared with GAN-based methods, our method demonstrated improvements of 5.5%, 5.9%, and 13.2% in the metrics of low-level pixel features, shallow structural features, and deep structural features, respectively.

Intelligent design of shear wall layout based on diffusion models

AbstractThis study explores artificial intelligence (AI) for shear wall layout design, aiming to overcome challenges in data feature sparsity and the complexity of drawing representations in existing AI‐based methods. We pioneer an innovative method leveraging the potential of diffusion models, establishing a suitable drawing representation, and examining the impact of various conditions. The proposed image‐prompt diffusion model incorporating a mask tensor featuring tailored training methods demonstrates superior feature extraction and design effectiveness. A comparative study reveals the advanced capabilities of the Struct‐Diffusion model in capturing engineering designs and optimizing performance metrics such as inter‐story drift ratio (in elastic analysis), offering significant improvements over previous methods and paving the way for future innovations in intelligent designs.

A numerical validation between the neutron transport and diffusion theories for a spatial kinetics problem

In this paper, a comparative analysis of numerical results of the neutron transport and diffusion theories for steady-state and transient multigroup problems is presented. The neutron transport equation is known as the one that best describes the behavior of the neutron population in a nuclear reactor. However, due to the difficulty of working with its complete form, other models are considered as approximations to this equation. One such approximation is the neutron diffusion equation, which uses the Fick's Law. It is well known, however, that the diffusion model may not work well under specific conditions. The objective of this work is to establish a quantitative comparison of numerical results obtained for the K dominant eigenvalue and for the scalar fluxes from the two theories and to analyze the influence of the model on the results.

Beta Brownian motion

Brownian motion (BM) is the paradigmatic model of diffusion. Transcending from diffusion to anomalous diffusion, the principle Gaussian generalizations of BM are Scaled BM (SBM) and Fractional BM (FBM). In the sub/super diffusivity regimes: SBM is characterized by aging/anti-aging, and FBM is characterized by anti-persistence/persistence. BM is neither aging/anti-aging, nor persistent/anti-persistent. Within the realm of diffusion, a recent Gaussian generalization of BM, Weird BM (WBM), was shown to display aging/anti-aging and persistence/anti-persistence. This paper introduces and explores the anomalous-diffusion counterpart of WBM—termed Beta BM (BBM) due to its inherent beta-function kernel structure—and shows that: the weird behaviors of WBM become even weirder when elevating to BBM. Indeed, BBM displays a rich assortment of anomalous behaviors, and an even richer assortment of combinations of anomalous behaviors. In particular, the BBM anomalous behaviors include aging/anti-aging and persistence/anti-persistence—which BBM displays in both the sub and super diffusivity regimes. So, anomalous behaviors that are unattainable by the prominent models of SBM and FBM are well attainable by the BBM model.

Viscoelastic nanofluids flow through Darcy–Forchheimer porous media with Newtonian conditions

This research is conducted by considering the Darcy–Forchheimer viscoelastic non-Newtonian fluids (NNF) generated by the movement of surface. Two types NNF namely Walter’s liquid B (WLB) and second grade (SG) fluids are chosen as the working fluids. Cattaneo–Christov energy diffusion model is carried out for the energy transmission analysis. Newtonian heating (NH) and Newtonian concentration (NC) conditions are imposed for the solutal and thermal analysis. The model is formulated by the application of boundary layer technique. These equations are converted into non dimensional model by the implication of similarity variables. The numeric solutions are executed by the utilization of Runge–Kutta–Fehlberg-45 order scheme. The results are presented by numeric benchmarks and graphical illustrations. A comparative discussion of WLB and SG fluids is reported. A comparative visualization shows that the higher velocity profile is achieved for SG fluid case while temperature and concentration profiles are larger in the case of SG fluid. The Brownian movement and thermophoretic constraints have similar influences on temperature. The Brownian movement constraint has reverse trends on concentration and temperature.

Crowdsourcing with the drift diffusion model of decision making.

Crowdsourcing involves the use of annotated labels with unknown reliability to estimate ground truth labels in datasets. A common task in crowdsourcing involves estimating reliabilities of annotators (such as through the sensitivities and specificities of annotators in the binary label setting). In the literature, beta or dirichlet distributions are typically imposed as priors on annotator reliability. In this study, we investigated the use of a neuroscientifically validated model of decision making, known as the drift-diffusion model, as a prior on the annotator labeling process. Two experiments were conducted on synthetically generated data with non-linear (sinusoidal) decision boundaries. Variational inference was used to predict ground truth labels and annotator related parameters. Our method performed similarly to a state-of-the-art technique (SVGPCR) in prediction of crowdsourced data labels and prediction through a crowdsourced-generated Gaussian process classifier. By relying on a neuroscientifically validated model of decision making to model annotator behavior, our technique opens the avenue of predicting neuroscientific biomarkers of annotators, expanding the scope of what may be learnt about annotators in crowdsourcing tasks.

High-temperature oxidation characteristics of 20CrMnTi electroplated with copper

During high-temperature forming and phase change heat treatment, the steel surface is susceptible to oxidation. This can negatively affect the surface quality of steel parts. The production of antioxidant coatings on the surface of steel parts is important for high-temperature precision forming and heat treatment. In this paper, the high-temperature oxidation tests of 20CrMnTi steel, copper, and 20CrMnTi electroplated with copper are conducted. The 20CrMnTi samples showed varying degrees of mass gain above 900 °C; the copper samples showed varying degrees of mass gain above 700 °C; and the process of mass gain for copper-plated 20CrMnTi followed the same trend as for copper during heating and holding. For heated samples of 20CrMnTi plated with copper, the copper coating was oxidized and residual copper was observed in the oxide layer. The distance between the residual copper and the surface of the sample is greater than between the residual copper and the junction of the plating layer and the substrate. Oxygen diffuses through internal defects in the material at a higher rate than copper oxidation. In this study, an oxygen diffusion model and an antioxidant protective layer model at high temperatures were suggested. The protective layer against oxidation recommended in this research is a dense coating consisting of single crystals or fine grains free of porosities and cracks.

Optical Control of Directional Exciton Migration in Porphyrin‐Based Metal−Organic Frameworks

AbstractPorphyrin‐based metal‐organic frameworks (MOFs) have received great attention for solar photochemistry applications. Manipulating the direction of energy (exciton) migration plays an essential role in boosting the light‐harvesting efficiency of porphyrin‐based MOFs. Here, based on the time‐dependent exciton‐exciton annihilation rate, an optical control of directional exciton migration is demonstrated in cobalt‐porphyrin surface‐supported MOF nanofilms (Co‐TCPP SURMOFs). Utilizing photocarrier densities or lattice temperature as the control knob, the dynamic form of exciton migration in Co‐TCPP SURMOFs can be switched between the nearest‐neighbor 1D interaction and the quasi‐isotropic 2D interaction, which is well explained using a diffusion model of exciton‐exciton scattering. This results provide an optical approach to regulating exciton migration in MOFs, which enhances understanding of the excitonic behavior in artificial light harvesters.

Hydrothermally Produced Activated Carbon Impregnated with ZnO for the Adsorptive Removal of Toxic Pharmaceutical Contaminants from Aqueous Solution

This research explores the adsorption (AD) of diclofenac sodium (DS) onto a Hydrothermally produced activated carbon impregnated with ZnO (HTC-AC/ZnO) surface, considering various factors such as initial concentration (IC), adsorbent dose, contact time, and pH. The characterization of HTC-AC/ZnO was performed using X-ray diffractometer (XRD), scanning electron microscope (SEM), Fourier-transform infrared spectroscopy (FTIR), and nitrogen physisorption spectroscopy (BET). Tests were conducted with different adsorbent doses (0.5–4 g/L) at 303 K and various initial diclofenac concentrations (ranging from 50 mg/L to 250 mg/L) to observe their effects. Additionally, pH values were altered from 2 to 12 to study their influence on AD. Kinetic studies, thermodynamic studies, and AD isotherm models were examined. The Temkin isotherm model (TIM) was found to be the most accurate for DS-AD on HTC-AC/ZnO. For DS-AD on HTC-AC/ZnO, pseudo-first-order models (PFOM), intraparticle diffusion model (IPDM), and pseudo-second-order models (PSOM) were applied, with a correlation coefficient of 0.945, indicating a good fit for PFOM. The kinetics suggested rapid adsorption. Notably, the HTC-AC/ZnO composite exhibited consistent AD characteristics across four consecutive cycles, with a removal efficiency exceeding 99.38%. This suggests that HTC-AC/ZnO is an appropriate and economically viable adsorbent for the elimination of DS from water-based solutions. The investigation provides compelling evidence that HTC-AC/ZnO is a viable adsorbent for the effective elimination of DS from water sources.