The content of flavonoids especially baicalin and baicalein determined the medical quality Scutellaria baicalensis which is a Chinese traditional medicinal plant. Here, we investigated mechanism responsible for composition in S. under water deficit condition. transcription levels several genes are involved flavonoid biosynthesis were stimulated by deficit. Under condition, fifteen up-regulated proteins, three down-regulated proteins other six detected proteomic analysis. identified include...

Sign language translation (SLT) is an important application to bridge the communication gap between deaf and hearing people. In recent years, research on SLT based neural frameworks has attracted wide attention. Despite progress, current still in initial stage. fact, systems perform poorly processing long sign sentences, which often involve long-distance dependencies require large resource consumption. To tackle this problem, we propose two explainable adaptations traditional models using...

Most superpixel methods are sensitive to noise and cannot control the number precisely. To solve these problems, in this paper, we propose a robust method called fuzzy simple linear iterative clustering (Fuzzy SLIC), which adopts local spatial C-means dynamic superpixels. We develop fast precise algorithm onion peeling (OP) algorithm. Fuzzy SLIC is insensitive most types of noise, including Gaussian, salt pepper, multiplicative noise. The OP can accurately without reducing much computational...

In most multiobjective optimization problems of electrical machines, the weighted function method is used to convert them into single-objective problems. This paper applies a kind new evolutionary algorithms (MOEAs), called adaptive black hole (AMOBH) algorithms, achieve effective tubular coreless linear permanent magnet synchronous motor (LPMSM). To reduce computation cost MOEAs, one-layer analytical model (AM) presented for LPMSM in this paper. The accuracy simplified AM verified by...

This research proposes using a neural network to detect and identify the landmark points of carapace Chinese mitten crab, with aim improving efficiency in observation, measurement, statistics breeding sales. A 37-point localization framework was developed for carapace, dataset augmented through random distortions, rotations, occlusions enhance generalization capability. Three types convolutional models were used compare detection accuracy, ability, model power consumption, different loss...

This paper proposes a new multiobjective evolutionary algorithm based on the black hole with individual density assessment (cell density), called “adaptive algorithm” (AMOBH). Cell has characteristics of low computational complexity and maintains good balance convergence diversity Pareto front. The framework AMOBH can be divided into three steps. Firstly, front is mapped to objective space parallel cell coordinate system. Then, adjust strategies adaptively, Shannon entropy employed estimate...

Superpixel-based fast fuzzy C-means clustering (SFFCM) is an efficient method for color image segmentation. However, it sensitive to noise and blur. Its superpixel called multiscale morphological gradient reconstruction (MMGR) time consuming. In this paper, we propose improved SFFCM (ISFFCM) which replaces the MMGR in with simple linear iterative (Fuzzy SLIC). Fuzzy SLIC faster more robust than most types of noise, including salt pepper Gaussian multiplicative noise. It also validation...

In this paper, we propose a new unsupervised attention-based cycle generative adversarial network to solve the problem of single-image dehazing. The proposed method adds an attention mechanism that can dehaze different areas on basis previous (GAN) dehazing method. This not only avoids need change haze-free area due overall style migration traditional GANs, but also pays degrees haze concentrations be changed, while retaining details original image. To more accurately and quickly label haze,...

Convolutional neural networks (CNNs) are usually used as a backbone to design methods in biomedical image segmentation. However, the limitation of receptive field and large number parameters limit performance these methods. In this paper, we propose graph network (GNN) based method named GNN-SEG for segmentation brain tissues. Different conventional CNN methods, takes superpixels basic processing units uses GNNs learn structure Besides, inspired by interaction mechanism biological vision...

Computational methods are often applied to identify essential proteins from protein-protein interaction networks. In this paper, inspected by node and edge clustering coefficient (NEC) PeC, we propose an improved version of (INEC) which both considers dual topological characteristics the network high false positives data. We apply it for identification proteins. And implement three versions INEC combine different biological information. order not be confused, call first one INEC0 dosen't...

In this letter, we propose a hierarchical segmentation (HS) method for color images, which can not only maintain the accuracy, but also ensure good speed. our method, HS adopts fuzzy simple linear iterative clustering (Fuzzy SLIC) to obtain an over-segmentation result. Then, uses fast C-means (FFCM) produce rough result based on superpixels. Finally, takes non-iterative K-means using priority queue (KPQ) refine validation experiments, tested and compared it with state-of-the-art image...

In this paper, we propose a method called superpixel tensor pooling tracker which can fuse multiple midlevel cues captured by superpixels into sparse pooled features. Our first adopts the to generate different patches (superpixels) from target template or candidates. Then for each superpixel, it encodes including HSI color, RGB and spatial coordinates histogram matrix construct new feature space. Next, these matrices are formed third order tensor. After that, is representation. incremental...

In high-risk industries such as safety production and aerial work, the use of helmets by workers constitutes a crucial measure for preventing injuries. Automated detection helmet-wearing can enhance objectivity reduce labor costs. However, current object algorithms may overlook small-sized targets, leading to missed detections. To address this issue, paper proposes an improved algorithm based on YOLOv5s. First, attention mechanism is introduced into multi-scale fusion module YOLOv5s...

This paper proposes an improved (TS-NSGA-II) algorithm based on non-dominated sorted genetic algorithm-II (NSGA-II) algorithm, which is used for multiple unmanned aerial vehicles (multi-UAVs) to perform cooperative detection tasks in complex environments with constraints. Firstly, under the consideration of various constraints, such as maximum flight distance, minimum safe distance and time points, a multi-objective optimization function established including profit, energy consumption...

Superpixel is widely used in image processing. And among the methods for superpixel generation, clustering-based have a high speed and good performance at same time. However, most are sensitive to noise. To solve these problems, this paper, we first analyze features of Then according statistical noise, propose novel centroid update approach enhance robustness methods. Besides, superpixel-based edge detection method. The experiments on BSD500 dataset show that our can significantly noisy...

Using graph theory to identify essential proteins is a hot topic at present. These methods are called network-based methods. However, the generalization ability of most not satisfactory. Hence, in this paper, we consider identification as multi-objective optimization problem and use novel method solve it. The result set Pareto solutions. Every solution vector which has certain number protein candidates considered an independent predictor or voter. We voting strategy assemble results these...

Identification of essential proteins is a hot topic in bioinformatics. In recent years, various traditional methods have been proposed, which usually take topological features to rank the and then set threshold for selecting proteins. Some researchers also tried machine learning or deep prediction However, these can not well extract protein-protein interaction (PPI) network. Besides, although some scholars proposed combine biological information with PPI network reduce noise data, how...

Superpixels can preserve the structure and reduce redundancy of original image. Because these advantages, superpixel generation or segmentation is widely used as a pre-processing step in many image processing tasks. Although superpixels be employed to computational complexity, some challenges, such non-Euclidean feature learning problem introduced by superpixels, still exist. This survey provides comprehensive overview state-of-the-art methods, major commonly evaluation metrics, applications...

The classifier is an essential tool for the development of contemporary engineering technology. application classifiers to categorize mixed-sized particles into multi-stage uniform particle sizes. In current studies, in obtain their initial velocity when feeding. classification effect impacted by inability precisely control state particles. To solve this problem, a pusher feed was designed study, and numerical simulation performed investigate its flow field characteristics performance using...

In this paper, a novel algorithm called non-local adaptive mean filter (NAMF) for removing salt-and-pepper (SAP) noise from corrupted images is presented. We employ an efficient window detector with size to detect the noise, noisy pixel will be replaced by combination of its neighboring pixels, and finally we use SAP based reconstruct intensity values pixels. Extensive experimental results demonstrate that NAMF can obtain better performance in terms quality restoring at all levels noise.

Task assignment is the main conundrum of multiple unmanned aerial vehicles (multi-UAVs) cooperative task execution, which has great research value. To solve problem scheduling for multi-UAVs, a method based on improved genetic algorithm proposed in this paper. Firstly, mathematical model established application environment multi-UAVs attacking targets. Secondly, aiming at slow convergence speed and easy to fall into local extremum algorithm, Metropolis criterion simulated annealing...