Abstract Background Influenza is an infectious respiratory disease that can cause serious public health hazard. Due to its huge threat the society, precise real-time forecasting of influenza outbreaks great value our public. Results In this paper, we propose a new deep neural network structure forecasts influenza-like illness rate (ILI%) in Guangzhou, China. Long short-term memory (LSTM) networks applied precisely forecast accurateness due long-term attribute and diversity epidemic data. We...

<abstract> The path planning of robot is great significance for the logistics industry, which helps to improve efficiency warehousing, sorting and distribution. On basis ant colony algorithm, multi step search strategy used instead single strategy, pheromone update mechanism redesigned, smoothing configured performance algorithm. experimental results show that improved algorithm proposed in this paper can plan a shorter optimal on 16 * grid storage site, length saved by 9.21%. </abstract>

Abstract Objective. P300s are one of the most studied event-related potentials (ERPs), which have been widely used for brain–computer interfaces (BCIs). Thus, fast and accurate recognition is an important issue BCI study. Recently, there emerges a lot novel classification algorithms P300-speller. Among them, discriminative canonical pattern matching (DCPM) has proven to work effectively, in spatial (DSP) filter can significantly enhance features P300s. However, ERPs space varies with time,...

To meet the requirements of some intelligent vehicle monitoring system, software integrates Global Position System (GPS), Geographic Information (GIS) and for Mobile communications (GSM) in whole. The structure, network topology, functions, main technical features their implementation principles system are introduced. Then hardware design terminal is given short. Communication process data transmission between server client (relay server) through TCP/IP UDP protocol discussed detail this...

BACKGROUND: Medical image security has been paid more attention in the medical field. OBJECTIVE: In order to achieve a higher level of encryption, this paper proposes stream cipher enhanced logic mapping encryption method. METHODS: According theory cipher, method uses Chebyshev map form key. A series coding operations are used set initial value before chaos processing. Combining with logical mapping, original information is encrypted by from X and Y dimensions. RESULTS: The experimental...

The pilot’s situational awareness (SA) mainly depends on the perception, understanding and prediction of information obtained from man-machine display interface, therefore there is an intrinsic link between interface design level SA. A quantitative SA model, which considers influence importance SA, characteristics human cognition Bayesian conditional probability theory, was proposed based attention allocation model built previously. This expected to be used predict levels under different...

Human motion capture techniques (MOCAP) are widely applied in many areas such as computer vision, animation, digital effect and virtual reality. Even with professional MOCAP system, the acquired data still always contains noise outliers, which highlights need for essential refinement methods. In recent years, approaches have been developed, including signal processing based methods, sparse coding methods low-rank matrix completion However, is a challenging task due to complexity diversity of...

Color, texture, and shape are the common features used for retrieval systems. However, many medical images have a spot of color information. Therefore, discriminative texture should be extracted to obtain satisfied result. In order increase credibility process, can combined image retrieval. Meanwhile, more require processing time, which will decrease speed. this paper, wavelet decomposition is adopted generate different resolution images. Bag-of-feature, LBP feature from three...

We consider the outer synchronization between drive-response systems on networks with time-varying delays, where we focus case when underlying are not strongly connected. A hierarchical method is proposed to characterize large-scale without strong connectedness. The algorithm can be implemented by some programs overcome difficulty resulting from scale of networks. This allows us obtain two different kinds sufficient criteria assumption being connected, combining theory asymptotically...

This paper studies the feature based face detection algorithm. Based on principal component analysis, vector space is extracted to construct weak classifier. Combined with Adaboost algorithm strong classifier, an for presented. The performance of tested MIT+CMU database, results show that in running time and accuracy significantly better than neural network support machine

<p indent="0mm">Claystone is one of the main host rocks containing rare earth elements (REE). It has a huge potential REE resource. Currently, due to unclear understanding occurrence state in claystone, development and utilization claystone are limited. In this study, REE-rich samples from Baozhuang area Wuan City, Hebei Province were selected for analyzing their petrological characteristics using rock mineral identification, automated identification characterization system (AMICS), X-ray...

As artificial intelligence technology continues to advance, the utilization of deep learning in sonar image target detection is growing. This trend has garnered significant interest maritime defense infrastructure and marine economic development as a promising research avenue. Nevertheless, existing methods face challenges due inherent characteristics images, such low resolution, blurred edges, speckle noise. These issues often result large model parameters slow inference speeds. To mitigate...

Image matting is an important technology for image synthesis and video editing. The existing method based on pixel pair optimization can significantly improve the accuracy of matting, but with increase resolution, difficulty increases. This paper proposes a group strategy, which transforms large-scale combinatorial problem into multiple small-scale problems. Based grouping we designed algorithm natural according to assumption that alpha within similar. In algorithm, color average intra-group...

Abstract An artificial neural network (ANN) is a mathematical or computational model that simulates the structure and function of biological networks used to evaluate approximate functions at given points. After developing training algorithm, resulting will be solve image recognition problems, control optimization, etc. In process ANN training, algorithm backpropagation in case convex optimization functions. The article analyzed test for experiments also study effect number layers on quality...

The paper analyzes the concept, elements and data structure of multi-mode transportation network, proposes a polycentric hierarchy description model integrated discusses construction method network using GIS technology. design system function modules provides visual edition platform to realize conversion vectorgraph data, as well functions storage, modification query attribute socioeconomic network. Through information update daily maintenance this can offer accurate complete for...

Abstract Face alignment and tracking play important roles in facial performance capture. Existing data‐driven methods for monocular videos suffer from large variations of pose expression. In this paper, we propose an efficient robust method task by introducing a novel supervised coordinate descent with 3D bilinear representation. Instead learning the mapping between whole parameters image features directly cascaded regression framework current methods, learn individual sets mappings...