Search

• Current methods cannot generate compound expressions precisely and variously. • Generating various compound expressions is formulated into an optimization problem. • We propose EvoGAN, which EAs, GANs and facial expression recognizers. • EvoGAN models and generates expression for every input figure. The image synthesis technique is relatively well established which can generate facial images that are indistinguishable even by human beings. However, all of these approaches uses gradients to condition the output, resulting in the outputting the same image with the same input. Also, they can only generate images with basic expression or mimic an expression instead of generating compound expression. In real life, however, human expressions are of great diversity and complexity. In this paper, we propose an evolutionary algorithm (EA) assisted GAN, named EvoGAN, to generate various compound expressions with any accurate target compound expression. EvoGAN uses an EA to search target results in the data distribution learned by GAN. Specifically, we use the Facial Action Coding System (FACS) as the encoding of an EA and use a pre-trained GAN to generate human facial images, and then use a pre-trained classifier to recognize the expression composition of the synthesized images as the fitness function to guide the search of the EA. Combined random searching algorithm, various images with the target expression can be easily sythesized. Quantitative and Qualitative results are presented on several compound expressions, and the experimental results demonstrate the feasibility and the potential of EvoGAN. The source code is available at https://github.com/ECNU-Cross-Innovation-Lab/EvoGAN. [ABSTRACT FROM AUTHOR]

In this paper, a dynamic multiobjective evolutionary algorithm (DMOEA) with an adaptive response mechanism selection strategy is proposed to address the shortcoming that a single response mechanism is suitable only for solving a certain type of dynamic multiobjective optimization problem. The proposed algorithm combines an adaptive response mechanism selection (ARMS) strategy and a multiobjective evolutionary algorithm based on decomposition (MOEA/D), and it is denoted as the MOEA/D-ARMS. Unlike the existing approaches, the ARMS strategy can adaptively select effective response mechanisms from the response mechanism pool based on the recent performance of each response mechanism. Four representative response mechanisms are selected to form the response mechanism pool. An overall evaluation strategy that assigns rewards to the response mechanism is adopted, and a probability-based method that is used to decide which response mechanism can be used to generate a new solution is employed. The proposed MOEA/D-ARMS algorithm is tested on two groups of test instances and compared with the decomposition-based and dominance-based DMOEAs. The results of the proposed MOEA/D-ARMS algorithm are superior to the compared algorithms, demonstrating its effectiveness. • This paper proposed an adaptive response mechanism selection (ARMS) framework and a dynamic MOEA. • The ARMS framework can adaptively select the most effective response mechanisms based on their recent performance. • An overall evaluation strategy that assigned rewards to the response mechanism was adopted. • The simple ARMS framework can easily incorporate more response mechanisms. • The statistical results clearly demonstrated that MOEA/D-ARMS were superior to the compared algorithms. [ABSTRACT FROM AUTHOR]

For the first time, a novel nanocomposite, zeolitic imidazolate framework/magnetic layered double hydroxide (Fe 3 O 4 @LDH/ZIF), was successfully fabricated through in situ synthetic method with Fe 3 O 4 as magnetic core and composite as shell. The resulting materials were characterized by XRD, TEM, FT-IR, VSM, TGA and nitrogen adsorption-desorption analysis. The Fe 3 O 4 @LDH/ZIF was employed as a sorbent in dispersive micro-solid phase extraction (D-μ-SPE) for enrichment of four endocrine disrupting compounds (EDCs) from milk samples prior to ultra-high performance liquid chromatography coupled with tandem mass spectrometry (UHPLC-MS/MS) analysis. The synthesized nanocomposite showed bigger specific surface area and better absorption capacity than Fe 3 O 4 @LDH. The equilibrium adsorption capacities (Qe, mg/g) of EDCs on the Fe 3 O 4 @LDH/ZIF reached up to 47.4–138.8 mg/g, mainly resulting from hydrogen bonding, π-π interaction and electrostatic interaction between EDCs and nanocomposite. Several variables affecting the extraction efficiency were investigated and optimized. The method displays low limits of detection (3–15 ng L−1), good linearity (10–5000 ng L−1), acceptable precision (RSD < 7.8%) and accuracy (RE < 5.2%). Satisfactory repeatability was obtained with RSD in the range of 1.2–7.8%. This work provides a promising approach for the development of new sorbent media in sample preparation for the improvement of analytical performance. [Display omitted] • A novel core-shell Fe 3 O 4 @LDH/ZIF adsorbent was successfully synthesized by in situ growth method. • The synthesized nanocomposite features large surface area, good magnetic responsibility and recyclability. • The Fe 3 O 4 @LDH/ZIF enriched four EDCs effectively in milk by multi-interactions. • Extraction parameters were optimized and enrichment mechanism was described. [ABSTRACT FROM AUTHOR]

In this paper, we investigated pollen assemblages of 92 lake center surface sediments of arid and semi-arid northern China to statistically disclose the pollen–vegetation relationships at the regional scale and explore considerations and limitations for lacustrine pollen-based reconstructions. As a result, the Principal component analysis (PCA), the Indicator species analysis, the Discriminate analysis (DA), the Random forest (RF), and the Supportive vector machine (SVM) generally reveal positive correlations between pollen assemblages and vegetation types or ecotones and suggests capabilities of pollen to represent regional vegetation and climate. However, the DA, RF, and SVM reveal that the samples of desert steppe, typical steppe, and meadow steppe cannot be differentiated by pollen assemblages, suggesting weaknesses to represent regional vegetation and climate. Mental test analyses of pollen assemblages and azonal vegetation of the dune desert and the Gobi desert yield better correlations than that of regional vegetation, suggesting that the pollen of azonal vegetation distort the regional pollen–vegetation relationships and thus weaken the capabilities to represent regional vegetation and climate. Our results present considerations and limitations for pollen-based past vegetation and climate reconstructions and provide references for interpretations of lacustrine fossil pollen records in arid and semi-arid northern China. • Lake surface sediment pollen–vegetation revealed general correlation relationships. • Azonal vegetation pollen bias regional pollen–vegetation relationships in desert. • Lake sediment pollen assemblages represent azonal vegetation better in desert. [ABSTRACT FROM AUTHOR]

Grain size is commonly used as a proxy to reconstruct past lake hydrology; however, published research rarely links grain size with modern processes, and uncertainties exist for interpretation. In this paper, we decompose the grain size frequency distributions (GSFD) for 196 lake surface sediments from 61 lakes and 9 reservoirs across the Inner Mongolia Plateau and adjacent areas and divide them into five GSFD components. We statistically analyze the relationship between the GSFD components and influencing variables to determine the taphonomic dynamics and their environmental implications. Within individual lakes, we found that GSFD component 4 (74.00–209.30 μm) and GSFD component 5 (209.30–592.00 μm) dominate near the lake margin and reduce toward the lake center in proportions. By contrast, GSFD component 2 (2.75–15.56 μm) increases toward the lake center. Therefore, the proportions of GSFD component 2, 4, and 5 reflect the relative distance to the lakeshore and allow inferences about lake area changes. We apply this large modern dataset to a sediment core obtained from Lake Wulanhushao (WLHS) on the Inner Mongolia Plateau and link the lake surface sediment-based GSFD components to BEMMA end-members of grain size records to define their environmental implications over the last 18.59 cal. kyrs BP. We found that the Lake WLHS has responded to the orbital and suborbital climate changes since the last deglaciation. Specifically, maximum lake transgression prevailed during the early to middle Holocene (11.70–7.10 cal. kyr BP), medium lake transgression dominated during 17.05–15.80 cal. kyr BP, and weak transgression developed during 14.60–12.90 cal. kyr BP. Maximum lake regressions prevailed before 17.05 cal. kyr BP, during 15.80–14.60 cal. kyr BP and 12.90–11.70 cal. kyr BP. Our reconstructed early Holocene maximum lake area is consistent with regional abiotic proxy-based moisture records, but contrasts with the regional biotic proxy-based moisture records, which reconstruct increasing moisture conditions. Our new method to determine lake area changes will provide valuable data to understand the climate dynamics of northern China. • We numerically revealed the environmental implications of grain size components of lake surface sediment. • Linking lake surface sediment grain size component to core end members provide basis for interpretations. • Maximum lake transgression prevailed during the early Holocene in northern China. • Abiotic proxies-based early Holocene lake transgression is distinct with biotic proxies' reconstructions. [ABSTRACT FROM AUTHOR]

Submarine pipelines that transport crude oil and natural gas are often in a complex marine geological environment and may become unstable and fail upon impact by submarine landslides. Previous research has mostly focused on the impact forces exerted by submarine landslides on suspended pipelines, but the impact of submarine landslides on pipelines laid on the seafloor at various impact angles, θ , have been relatively infrequently discussed, and the effects of suspended height, H , on the impact forces exerted by submarine landslides on pipelines have not been thoroughly investigated. In this study, based on the Herschel–Bulkley model, the impact forces exerted by a submarine landslide on laid-on or suspended pipelines at various impact angles θ were simulated using the computational fluid dynamics (CFD) approach. Equations for calculating the axial and normal drag coefficients of a submarine pipeline were proposed. The CFD numerical simulation results were rearranged based on the soil mechanics approach. By comparing the parameters, an essentially corresponding relationship was found between the soil mechanics and CFD approaches when the equations were used to calculate the impact forces exerted by a submarine landslide on a pipeline. In addition, a semi-analytical expression for the failure envelope was provided. Furthermore, the effects of H on the forces on a pipeline were discussed, and an equation for calculating the acting forces on a pipeline along the flow direction of a submarine landslide that comprehensively accounts for the effects of θ and H was proposed. The lift force was discussed preliminarily and the results provide a basis for further investigation. The achievement of this study is applicable for selecting locations of submarine pipeline routes and for designing submarine pipelines. [ABSTRACT FROM AUTHOR]