Artificial intelligence generated content (AIGC) has received significant attention at present. As the numerous generative models proposed, the emerging diffusion model has attracted extensive attention due to its highly interpretable mathematical properties and the ability to generate high-quality and diverse results. Nowadays, diffusion models have achieved remarkable results in the field of condition-guided image generation. This achievement promotes the development of diffusion models in other conditional tasks and has various applications in areas such as movies, games, paintings, and virtual reality. For instance, the diffusion model can generate high-resolution images in text-guided image generation tasks while ensuring the quality of the generated images. In this paper, we first introduce the definition and background of diffusion models. Then, we present a review of the development history and latest progress of conditional image generation based on diffusion models. Finally, we conclude this survey with discussions on challenges and future research directions of diffusion models.

The rotated flux method could be used to effectively eliminate the shock instability of the approximate Riemann solver and suppress the generation of non-physical phenomenon if the equations to be solved meet the rotational invariance. For the 2D ideal magnetohydrodynamics (MHD) and shallow water magnetohydrodynamics (SWMHD) equations, the rotation-like matrix of flux function was constructed, and the corresponding rotational invariance theorem was given with proof, which was then used to deal with the governing equations applying quasi-1D method to derive the semi-discrete rotated flux scheme. Combining entropy stable flux and anti-diffusive flux by a flux limiter, a new flux that can adaptively adjust the dissipation term was obtained. Numerical experiments show that the new scheme can accurately capture the structure of solution, has high resolution, strong robustness and can be easily extended to higher dimensions.

High spatial resolution remote sensing images contain rich information, it is therefore very important to study their semantic segmentation. Traditional machine learning methods appear low accuracy and efficiency when used for segmenting high-resolution remote sensing images. In recent years, the deep learning method has developed rapidly and has become the mainstream method of image semantic segmentation. Some scholars have introduced SegNet, Deeplabv3+, U-Net and other neural networks into remote sensing image semantic segmentation, but these networks have only limited effect in remote sensing image semantic segmentation. This paper improves the U-Net network for semantic segmentation of remote sensing images. Firstly, an improved convolutional attention module channel interaction and spatial group attention module (CISGAM) is embedded in the feature extraction stage of the U-Net network, so that the network can obtain more effective features; secondly, a residual module is used in the decoding layer to replace the ordinary convolutional layer to avoid the degradation of the model. In addition, we use an attention pyramid pooling module (APPM) with CISGAM to connect the encoder and decoder of U-Net to enhance the network's extraction of multi-scale features. Finally, experiments are carried out on the UC Merced dataset with 0.3m resolution and the GID dataset with 1m resolution. Compared with the original networks such as U-Net and Deeplabv3+, the mean intersection over union （MIoU) of our method on the UCM dataset has increased by 14.56% and 8.72%, and the mean pixel accuracy （MPA) has increased by 12.71% and 8.24%, respectively. In the classification results on the GID dataset, the classification accuracy of waters, buildings and other objects has also been greatly improved. Compared with the original CBAM and PPM, the CISGAM and APPM also achieve certain performance improvement. The experimental results show that the feasibility and robustness of the model is stronger than traditional networks, and it can improve the accuracy of semantic segmentation of high-resolution remote sensing images through stronger feature extraction capabilities, hence providing a new approach for intelligent interpretation of high-resolution remote sensing images.

Two simple graphs are commutative if there exists a labelling of their vertices such that their adjacency matrices can commute. This paper gives three necessary conditions ensuring the commutativity of certain graphs from Perron vectors, the number of main eigenvalues, the regularity of graphs. Then we construct new commutative graphs by graph Kronecker product, Cartesian product and circulant matrix. Finally, for two commutative graphs, we provide two algorithms that can express one adjacency matrix as the matrix polynomial of another adjacency matrix with distinct eigenvalues, and compare their merits. Commutative graphs sharing common eigenvectors are essential to the study of spectral graph theory.

The characteristics of sea surface temperature distribution in the estuarine and offshore areas are of great significance for understanding natural processes such as ocean thermal and dynamic processes and air sea interaction,as well as the comprehensive results under the influence of human activities.Satellite data reconstruction is an important means to obtain dynamic and accurate sea surface temperature data.This study adopted a dynamic mode decomposition (DMD) for nonlinear systems with sufficient data,It took MODIS\SST data from January 2003 to July 2016,after eliminating the abnormal data and filling in the blank data,the ocean surface temperature data of the Yangtze River Estuary from August 2016 to December 2019 were reconstructed,and the reconstruction effect was evaluated.The research shows that DMD algorithm can solve the sampling problem of dynamic system well provided that the time series data is sufficient.Our results show that DMD algorithm combined with orthogonal right triangular (QR) decomposition can effectively reconstruct the sea surface temperature data of the Yangtze River Estuary,and the average root mean square error (RMSE) is 0.007 6.Further analysis shows that both DMD algorithm and DMD algorithm combined with QR decomposition have high precision.

With the rapid increase of data size, the defects of traditional spatial indexing become more and more apparent. In comparison, learning indexing is based on data distribution. Its volume will not expand with the increase of the amount of data, and can achieve better performance without performing hierarchical comparison. Nevertheless, there are still two difficulties in applying the idea of learning indexing to spatial data: (1) How to choose appropriate dimension reduction method to sort the spatial data. (2) How to simplify data distribution of the dimension reduced data and make it easy to fit. This paper proposes a new type of grid mixed cluster partition learning indexing (grid-ml) based on the idea of learning indexing. In view of the above two difficulties, grid-ml uses z curve to reduce the dimension, and deals with the jumping problem with double-layer grid structure. Then, the improved K-means clustering method is used to simplify data distribution. The results show that grid-ml builds fast with small spatial storage volume, and can query fast as well, demonstrating significant advantages over the traditional spatial indexing approach.

The Rio del Rey basin, located in the eastern margin of the Niger Delta Basin in West Africa, develops large-scale mud diapirs. There are some disputes about the activity time and mechanism of the diapirs. Based on the detailed interpretation and mapping of the 3D seismic profile of the Rio del Rey basin, the distribution characteristics of the mud diapir on the top surface of the Miocene Nguti structural layer are redefined. The results show that: (1) mud diapir on the top of the Nguti structural layer can be divided into the northern buried area, the central pierced area, and the southern pierced area; (2) the mud diapir in the northern buried area was formed before the deposition of the Nguti structural layer, and was a passive mud diapir controlled by early sedimentation. The mud diapir in the central pierced area was formed before the deposition of the Pliocene S0 structural layer, and was an active mud diapir developed under the action of gravity sliding structures. The mud diapir in the southern pierced area was formed after the deposition of the Nguti structural layer, and was a thrust mud diapir controlled by thrust structures at the front edge of the gravity structure; (3) the mud diapir as a whole is a tongue shaped body protruding southward, divided into two segments: the west segment along a NNW-NW direction, and the east segment along a NNE-NE direction.

Herein, a novel one pot synthesis of chiral δ-substituent caprolactam strategy which applies 1-substituent homoallylic amine as the starting material is reported. This method combines the enzymatic dynamic kinetic resolution (DKR) and ring cross metathesis. Moreover, the configuration of caprolactam can be controlled by using lipase and protease as the DKR catalyst. The enantiomeric excess value of obtained R- and S-caprolactam can reach up to 90% and 82%, the yield can also be above 82% and 70%.

In the new era of fintech, digital technology is the core driving force of the future development of financial industry. With the new technology and the risk of financial fraud escalating, fraud detection based on digital technology has become a new research hot spot. Meantime the research direction of financial fraud detection technology has shifted from traditional methods of improving expert experience and optimizing machine learning models to exploring graph machine learning methods for social network. This article focuses on social network, based on the development process of network analysis, from different perspectives of detecting abnormal individuals, suspicious groups and unhealthy intermediaries, with different technical methods of digital financial fraud detection as the main line, the existing social-oriented relational network fraud identification methods are investigated, and the future research trends and directions of digital financial fraud detection technologies are highlighted.

In this paper, the phenomenon of phase transition is studied in an asymmetric tristable model driven by Lévy noise and Gaussian white noise. First, the fourth-order Runge-Kutta algorithm is used to simulate the stationary probability density of the system. Then we observe the shape of the stationary probability density curve by adjusting the system parameters and noise parameters. It has been found that asymmetric parameter, additive noise intensity, multiplicative noise intensity, the stability index, and the skewness parameter can all induce phase transition. Moreover, the influence of additive noise intensity and multiplicative noise intensity on the number and height of stationary probability density peaks is opposite. We also found that under the same asymmetric parameters, the influence of positive and negative skewness parameters on the stationary probability density is different.

In this study, we analyzed the differences in the effects of different forest vacation activities and personality traits on participants' behavioral intentions through two quasi-natural experiments (experimental group-forest therapy and control group-forest tourism), and found that:(1) personality traits of participants in two groups of forest vacation activities were categorized as resilient, overcontroler, and emotional;(2) there were significant differences in the results of behavioral intentions among the different class of personality traits, and the overall distribution of resilient > overcontroler > emotional type groups was shown; (3) under the interaction effect of the two groups of forest vacation activities and the three personality traits, the results showed that the behavioral intention was influenced by both the class of personality traits and the forest vacation activities synergistic effects. The findings of this study expand the study of the subjective influences on behavioral intentions of forest therapists at the theoretical level, and at the practical level, they help to deepen the forest therapy products and provide empirical evidence for the rational arrangement of experiential contents.

The Yangtze River basin plays an important role in Chinese water resources allocation. What proves common knowledge is that it is particularly important to predict the water quality in the Yangtze River basin. Based on the existing research, the recurrent neural network (RNN) model with gate recurrent unit (GRU) and fully connected neural network (FCNN) are combined in this study to improve a multiple water quality parameter prediction (MWQPP) model. It is proposed to predict the four water quality parameters, such as pH, dissolved oxygen (DO), permanganate index (CODMn) and ammonia nitrogen (NH₃-N) in the Yangtze River basin. Based on 7 566 raw data of 23 water quality monitoring points in the Yangtze River basin from 2011 to 2018, the comparative experiments show that the root mean square error (RMSE), mean absolute error (MAE), mean absolute percentage error (MAPE) and coefficient of determination (R²) obtained from the MWQPP model's prediction results are better than traditional models, such as the multiple linear regression model, the random forest model, FCNN model and LSTM model, and the MWQPP model also has better robustness than these traditional water quality prediction models. As we can say, the MWQPP model can provide scientific, reasonable and effective support for water quality assurance and water management in Yangtze River basin.

The purpose of this paper is to establish some new oscillation criteria for the second-order nonlinear neutral differential equations with damping terms which have the form [a(t){\left|z^{\text{'}}(t)\right|}^{\alpha -\mathrm{1}}{z}^{\text{'}}{(t)]}^{\text{'}}+b(t){\left|z^{\text{'}}(t)\right|}^{\alpha -\mathrm{1}}{z}^{\text{'}}(t)+q(t){\left|x(\sigma (t))\right|}^{\beta -\mathrm{1}}x(\sigma (t))=\mathrm{0}， where z(t)=x(t)+p(t)x(\tau (t)). Our theorems improve, extent and unify a number of related results reported in the literature. The wide application of the obtained criteria is illustrated via examples.

Road traffic accidents occur frequently and have significant impacts on life, property, and society. However, existing researches on road traffic accident risk pay little attention on establishing an effective road network model that accurately describes the transmission characteristics of traffic accident risk. As a result, the accuracy of risk evaluation is limited. To address this issue, we propose a network geographically weighted regression method based on arrival time cost. We conduct experiments using data from roads, traffic violations, traffic accidents, and urban points of interest (POIs) in a city from 2018 to 2020. The experimental results demonstrate that the network geographically weighted regression method, based on arrival time cost, incorporates the propagation nature of traffic accident risk on the road. It significantly reduces evaluation errors and effectively evaluates road traffic accident risk and its influencing factors. The downtown area of the city exhibits high accident risk, primarily concentrated at intersections with heavy traffic flow and certain road points with inadequate transportation facilities. The impact of different types of road traffic violations and urban POIs on the risk of road traffic accidents varies significantly and exhibits strong spatial heterogeneity.

Similar to the influence on nilpotent residual of the totally permutability of subgroups,naturally,we can study their influence on σ-nilpotent residual when nilpotent groups are generalized to σ-nilpotent groups.The intersection of all normal subgroups N of G such that G/N is σ-nilpotent is called σ-nilpotent residual of G,and is denoted as G^{{\mathit{??}}_{\sigma }}. Let G=AB, where A and B are totally permutable subgroups of G,this paper gives some new conclusions that B normalises A^{{\mathit{??}}_{\sigma }} and centralises A^{{\mathit{??}}_{\sigma }} by using the concepts and theories of σ-nilpotent subgroups and σ-supersoluble subgroups,and by applying some properties and methods of complete Hall σ-set and finite group theory.

Nowadays, the applications of engineering-procurement-construction (EPC) general contracting mode in hydraulic engineering projects are increasing. And it will benefit EPC general contractor in value management by exploring the value-added mechanism. Firstly, this paper studies the value chain of hydraulic engineering, and identifies the value-added factors through literature research and expert interviews. Then the approach of questionnaire is applied to investigate the influence degree among each factor, and their centrality and cause degree are obtained by analysis method of DEMATEL. On the above basis, this paper divides the factors into 6 layers by adopting the analysis method of interpretative structural modeling (ISM), hence builds the value-added path of hydraulic engineering project with EPC general contracting mode. The result will provide the inspiration for EPC enterprises to formulate strategies in value management and further research.

Taking 30 provincal panel data from 1998 to 2013 of foreign direct investment (FDI) and inbound business tourism (IBT) in Chinese Mainland, this paper studied the spatial mismatch phenomena,characteristics, and dynamic mechanism of them based on the gravity model and the spatial mismatch index model. Research showed that spatial mismatch existed between FDI and IBT, and the spatial mismatch distance S expanded year by year. By the forecasting, S will reach 702.42 km in 2018. There were spatial differences in the spatial mismatch of FDI and IBT, showing some differences in the three areas and 30 provinces. The spatial mismatch in the East is more significant than the Central and the West. Guangdong and Jiangsu provinces ranked the top two while Gansu and Jilin provinces ranked the last two regarding the value of SMI. The spatial mismatch pattern in the East determined the spatial mismatch pattern of the whole country. Based on the development trend of FDI and IBT, the dynamic mechanism of the spatial mismatch of them was further analyzed. The characteristics of IBT and FDI are the inner driving powers of the spatial mismatch of them; The rapid developments of traffic are the direct powers; The regional differences between the East and the West are the external powers; Government policy supports are assistant powers.

Marching cubes (MC) is a classic algorithm for isosurface extraction. However, it can only be used to reconstruct closed surfaces, as it requires dividing the space into inside and outside. To solve this problem, a neural marching cubes algorithm for open surfaces is proposed. The key to our method is to introduce a new sort of point position called irrelative point. Since no isosurface exists between irrelevant points and inside or outside points, thus connecting the inside and outside of the target shapes. The determination of irrelevant points does not require complex networks or calculations, and can be directly determined by the distance from points to the surface. Meanwhile, a residual module introducing attention mechanism is adopted to replace the original network. In addition, new tessellations are designed, and open surfaces can be reconstructed with the help of irrelative points. Finally, a smoothing process is incorporated to improve the reconstruction quality of the border. By testing on both closed and open surfaces under various metrics, our experiments show that the proposed method achieves high-quality reconstruction of open surfaces while maintaining the capability of reconstructing closed surfaces.