BACKGROUND:Brain tumor segmentation plays an important role in assisting diagnosis of disease, treatment plan planning, and surgical navigation. OBJECTIVE:This study aims to improve the accuracy boundary using multi-scale U-Net network. METHODS:In this study, a novel with dilated convolution (DCU-Net) structure is proposed for brain based on classic structure. First, MR images are pre-processed alleviate class imbalance problem by reducing input background pixels. Then, spatial pyramid...

Insects can cause a major loss in stored grain, and early identification monitoring of insects become necessary for applying corrective action. Considering the effectiveness practicability, an image processing approach corresponding application embedded smartphones are proposed to identify count insects. For insect images acquired by mobile phones, one sliding window-based binarization is adopted release their nonuniform brightness, then connected domain-based histogram statistics presented...

Abstract Nonrigid registration of medical images is especially critical in clinical treatment. Mutual information a popular similarity measure for image registration; however, only the intensity statistical characteristics global consistency are considered MI, and spatial ignored. In this paper, novel intensity‐based combining normalized mutual with nonrigid proposed. The different parameters Gaussian filtering defined according to regional variance, adaptive introduced into local structure...

The purpose of representation learning is to encode the entities and relations in a knowledge graph as low-dimensional real-valued vectors through machine technology. Traditional methods like TransE, method which models relationships by interpreting them translations operating on embeddings graph’s entities, are effective for bases, but struggle effectively model complex one-to-many, many-to-one, many-to-many. To overcome above issues, we introduce new representation, reasoning, completion...

Human lying is influenced by cognitive neural mechanisms in the brain, and conducting research on lie detection speech can help to reveal of human brain. Inappropriate deception features easily lead dimension disaster make generalization ability widely used semi-supervised model worse. Because this, this paper proposes a algorithm combining acoustic statistical time-frequency two-dimensional features. Firstly, hybrid network based autoencoder (AE) mean-teacher established. Secondly, static...

Stemming from the object overlap and undertraining few samples, road dense detection is confronted with poor identification performance inability to recognize edge objects. Based on this, one transfer learning-based YOLOv3 approach for identifying objects in has been proposed. Firstly, Darknet-53 network structure adopted obtain pre-trained model, then training introduced as output layer special dataset of 2000 images containing vehicles; proposed random function adapted intialize optimize...

Abstract In order to degrade the inter-user interference caused by same beam selected for different users in mmWave massive MIMO systems, this paper proposes a joint selection combining cuckoo search (CS) and ant colony optimization (ACO) (referred as CSACO). Differently from existing interference-aware selection, candidate set (CBS) all is created according power distribution of beamspace channel, thereby can be classified into non-interfering (NIUs) interfering (IUs), NIUs will assigned...

Anticancer peptides (ACPs) have promising prospects for cancer treatment. Traditional ACP identification experiments the limitations of low efficiency and high cost. In recent years, data-driven deep learning techniques shown significant potential prediction. However, prediction models rely heavily on extensive training data. Furthermore, current publicly accessible dataset is limited in size, leading to inadequate model generalization. While data augmentation effectively expands existing...

In order to improve the long-term prediction accuracy of feed grain demand, a dynamic forecast model demand is realized with joint multivariate regression model, which correlation between and its influence factors analyzed firstly; then change trend various that affect predicted by using ARIMA model. The simulation results show proposed combined forecasting obviously higher than grey system Thus, it indicates algorithm effective.

The Editor-in-Chief and the publisher have retracted this article, which was submitted as part of a guest-edited special section. An investigation uncovered evidence systematic manipulation publication process, including compromised peer review. Editor no longer confidence in results conclusions article.CZ, JL, FZ either did not respond directly or could be reached.

The performance of TDOA positioning based on UWB is limited by the hyperbolic characteristics TDOA, especially for tags away from asymptote. Aiming at this problem, a new indoor system proposed. Firstly, TOF ranging adopted to build equations; then weighted centroid algorithm four base stations presented compute initial rough position tag; and residual weighting introduced optimize tag position; then, corresponding nonlinear equations, which will be algebraically transformed one distribution...

In this paper, we present the asymptotic properties of maximum quasi-likelihood estimators (MQLEs) in generalized linear models with adaptive designs under some mild regular conditions. The existence MQLEs equation is discussed. rate convergence and normality are also established. results illustrated by Monte-Carlo simulations.