Abstract Predictive maintenance (PdM) is currently the most cost-effective method for industrial equipment, offering improved safety and availability of mechanical assets. A crucial component PdM remaining useful life (RUL) prediction machines, which has garnered increasing attention. With rapid advancements in internet things artificial intelligence technologies, RUL methods, particularly those based on pattern recognition (PR) technology, have made significant progress. However, a...

The sample imbalance of expression datasets always leads to poor recognition results for minority classes. To solve this problem, we propose a facial network, called Residual Attentive Sharing Network (RASN). There is fact that different expressions have affinity features, which makes it possible the classes benefit from majority in feature extraction process, sharing features module compensate inadequate learning by features. In addition, an attention added highlight expression-related and...

In the research of robot systems, path planning and obstacle avoidance are important directions, especially in unknown dynamic environments where flexibility rapid decision makings required. this paper, a state attention network (SAN) was developed to extract features represent interaction between an intelligent its obstacles. An auxiliary actor discriminator (AAD) calculate probability collision. Goal-directed gap-based navigation strategies were proposed guide robotic exploration. The...

A watershed color image segmentation algorithm by integrating anisotropic diffusion with morphological reconstruction and automatic seeded region growing is proposed in this paper, the used for smoothing, which can maintain edges noises eliminated, as a result, watershed¡¯s over-segmentation decreased effectively. And then, an based on regions similarity of mean hue value saturation result developed. In order to reduce number small regions, removing also designed. The experimental results...

Moment-based features have been intensively explored in visual servoing (VS) applications, especially when lack of representational rigid points like the bottle production line. Convergence and many other constraints exist VS with image moments seldom simultaneously considered literature. This paper takes bottle-shape cylinders as a target proposes set new moment-based for global convergence application. Building on to desired location, we plan straight path from current camera frame one...

This article addresses the utilization of peripheral image correspondences obtained in cameras obeying unified model for path-planning-based visual servoing applications. The proposed approach consists reprojecting available onto an oriented virtual plane, which is designed to carry as many reprojections possible from correspondences, especially those periphery region. To realize this, a key transformation matrix and applied spherical projections correspondences. After this transformation,...

Combined with pattern recognition and information visualization technology, this paper proposes a visual classification method based on KPCA parallel coordinate plot KPPCP. This maps the raw data space into high-dimensional feature by means of nuclear function, then is deal PCA, finally processed visualized in coordinates. The experiment show that it can effectively extract non-linear features from , enlarge differences between various categories, provide Interactive visualization, enhance...

Modeling password distributions is a fundamental problem in security, benefiting the research and applications on guessing, strength meters, honey vaults, etc. As one of best segment-based models, WordPCFG has been proposed to capture individual semantic segments (called words) passwords. However, we find does not address well ambiguity segmentation by maximum matching, leading unreasonable many further inaccuracy modeling distributions. To ambiguity, improve probability with A*-like pruning...

In the image super-resolution algorithm model, a large receptive field can provide more valuable features, so Transformer with strong information interaction ability has achieved excellent results in processing applications. However, when range of reaches certain critical value, restoration performance also which indicates that unconditionally increasing will not continue to promote improvement performance. At same time, larger range, data model needs process, seriously increases...