Search results

The purpose of this research is to achieve multi-task autonomous driving by adjusting the network architecture of the model. Meanwhile, after achieving multi-task autonomous driving, the authors found that the trained neural network model performs poorly in untrained scenarios. Therefore, the authors proposed to improve the transfer efficiency of the model for new scenarios through transfer learning.

First, the authors achieved multi-task autonomous driving by training a model combining convolutional neural network and different structured long short-term memory (LSTM) layers. Second, the authors achieved fast transfer of neural network models in new scenarios by cross-model transfer learning. Finally, the authors combined data collection and data labeling to improve the efficiency of deep learning. Furthermore, the authors verified that the model has good robustness through light and shadow test.

This research achieved road tracking, real-time acceleration–deceleration, obstacle avoidance and left/right sign recognition. The model proposed by the authors (UniBiCLSTM) outperforms the existing models tested with model cars in terms of autonomous driving performance. Furthermore, the CMTL-UniBiCL-RL model trained by the authors through cross-model transfer learning improves the efficiency of model adaptation to new scenarios. Meanwhile, this research proposed an automatic data annotation method, which can save 1/4 of the time for deep learning.

This research provided novel solutions in the achievement of multi-task autonomous driving and neural network model scenario for transfer learning. The experiment was achieved on a single camera with an embedded chip and a scale model car, which is expected to simplify the hardware for autonomous driving.

Suggests that misunderstandings frequently occur when trying to understand Chinese language and culture, and so gives the implied meaning of various Chinese expressions and sayings such as greetings, thanks, respect, age, congratulations and taboo subjects.

The objective of the study is to identify and analyse drivers of smart manufacturing using integrated grey-based approaches. The analysis facilitates industry practitioners in the identification of preference of drivers through which smart manufacturing can be implemented. These drivers are explored based on existing literature and expert opinion.

Modern manufacturing firms have been adopting smart manufacturing concepts to sustain in the global competitive landscape. Smart manufacturing incorporates integrated technologies with a flexible workforce to interlink the cyber and physical world. In order to facilitate the effective deployment of smart manufacturing, key drivers need to be analysed. This article presents a study in which 25 drivers of smart manufacturing and 8 criteria are analysed. Integrated grey Technique for Order Preference by Similarity to Ideal Solution (grey TOPSIS) is applied to rank the drivers. The derived ranking is validated using “Complex Proportional Assessment – Grey” (COPRAS-G) approach.

In total, 25 drivers with 8 criteria are being considered and an integrated grey TOPSIS approach is applied. The ranking order of drivers is obtained and further sensitivity analysis is also done.

In the present study, 25 drivers of smart manufacturing are analysed. In the future, additional drivers could be considered.

The study presented has been done with inputs from industry experts, and hence the inferences have practical relevance. Industry practitioners need to focus on these drivers in order to implement smart manufacturing in industry.

The analysis of drivers of smart manufacturing is the original contribution of the authors.

The paper aims to present an approach for services in the domain of tourism based on a software application in the area of ontology engineering, showing a methodology for intelligent knowledge‐based P2P networks creation, in the tourism knowledge domain, given that, potential tourists share and organize their experiences, interests and knowledge. Using the proposed software application, they automatically exchange their knowledge, with an intelligent and transparent way, with other users that have the same or similar interests and make use of it.

The approach followed was categorizing tourism‐related interests and services into ontologies (system and user), then comparing them, using intelligent algorithms, suggesting new, unknown to the user, interests. The data were evaluated by experts in order to provide a guide for correct (according to the expert) interestingness of profile concepts.

The paper presents the outcomes of the software used, running on mobile devices, showing the connection for them, among user knowledge profiles and tourism services. It has been found that the return results (concepts) are of high interestingness to the user.

Experiments have been performed with one central ontology, used as reference and two user ontologies at the same time. The authors would suggest more experimentation, with more users connected concurrently.

Of high practical importance is the creation of intelligent social networking processes and user communities based on interests for further e‐commerce activities, based on a semantic framework.

The paper fulfils the need for intelligent interaction and distribution of knowledge and content to users in an autonomous way, anywhere.

Pressure-based methods have been demonstrated to be powerful for solving many practical problems in engineering. In many pressure-based methods, inner iterative processes are proposed to get efficient solutions. However, the number of inner iterations is set empirically and kept fixed during the whole computation for different problems, which is overestimated in some computations but underestimated in other computations. This paper aims to develop an algorithm with adaptive inner iteration processes for steady and unsteady incompressible flows.

In this work, with the use of two different criteria in two inner iterative processes, a mechanism is proposed to control inner iteration processes to make the number of inner iterations vary during computing according to different problems. By doing so, adaptive inner iteration processes can be achieved.

The adaptive inner iterative algorithm is verified to be valid by solving classic steady and unsteady incompressible problems. Results show that the adaptive inner iteration algorithm works more efficient than the fixed inner iteration one.

The algorithm with adaptive inner iteration processes is first proposed in this paper. As the mechanism for controlling inner iteration processes is based on physical meaning and the feature of iterative calculations, it can be used in any methods where there exist inner iteration processes. It is not limited for incompressible flows. The performance of the adaptive inner iteration processes in compressible flows is conducted in a further study.

Using equity theory, this study aims to examine the role of customer incivility in effecting service sabotage among hotel employees by recognizing the mediating role of revenge motivation and the moderating effect of emotion regulation.

A multi-wave, multi-source questionnaire survey was conducted with 291 employee–supervisor dyads at chain hotels in Shenzhen, China. Previously developed and validated measures for customer incivility, revenge motivation, emotion regulation and service sabotage were adopted to test the hypotheses.

Customer incivility increased employees’ revenge motivation and service sabotage. Emotion regulation acted as a boundary condition for customer incivility’s direct effect on revenge motivation and its indirect effect on service sabotage through revenge motivation. Cognitive reappraisal mitigated the detrimental influence of customer incivility, whereas expressive suppression worsened its adverse effects.

Managers should monitor and deter the emergence of uncivil behaviors, provide psychological support for employees experiencing customer incivility and encourage these employees to use cognitive reappraisal rather than expressive suppression as an emotion regulation strategy.

To the authors’ knowledge, no prior research has investigated the customer incivility–service sabotage relationship in the hotel industry. This study sheds light on how customer incivility can motivate service sabotage among hotel employees. Furthermore, the authors used equity theory rather than the commonly adopted resources perspective to offer new insights into the customer incivility–service sabotage relationship.

Automatic interpretation of Nominal Compounds is a crucial issue for many applications, for example, sentence understanding, machine translation, question-answering system and so forth. Many automatic interpretation models of Nominal Compounds use the strategies based on verbs or rules to obtain the interpretation of compounds. However, the performances of these models are still limited. The purpose of this paper is to propose an effective approach for automatic interpretation of Chinese nominal compounds.

The authors propose a top-down and bottom-up model based on rules and large-scale corpus for automatic interpretation of Nominal Compounds.

Experimental results demonstrate that the proposed model outperforms the state-of-the-art automatic interpretation model.

The paper is an up-to-date study of automatic interpretation for Nominal Compounds. It can help people understand the meaning of Nominal Compounds in reading. With a better understanding of Nominal Compounds, we can discover more hidden knowledge in them.

This study aims to fabricate and manipulate ensemble spin of negative nitrogen-vacancy (NV⁻) centres optimally for future solid atomic magnetometers/gyroscope. Parameters for sample preparation most related to magnetometers/gyroscope are, in particular, the concentration and homogeneity of the NV⁻ centres, the parameters’ microwave antenna of resonance frequency and the strength of the microwave on NV⁻ centres. Besides, the abundance of other impurities such as neutral NV centres (NV⁰) and substitutional nitrogen in the lattice also plays a critical role in magnetic sensing.

The authors succeeded in fabricating the assembly of NV centres in diamond and they determined its concentration of (2-3) × 10¹⁶ cm⁻³ with irradiation followed by annealing under a high temperature condition. They explored a novel magnetic resonance approach to detect the weak magnetic fields that takes advantage of the solid-state electron ensemble spin of NV⁻ centres in diamond. In particular, the authors set up a magnetic sensor on the basis of the assembly of NV centres. They succeeded in fabricating the assembly of NV centres in diamond and determined its concentration. They also clarified the magnetic field intensity measured at different positions along the antenna with different lengths, and they found the optimal position where the signal of the magnetic field reaches the maximum.

The authors mainly reported preparation, initialization, manipulation and measurement of the ensemble spin of the NV centres in diamond using optical excitation and microwave radiation methods with variation of the external magnetic field. They determined the optimal parameters of irradiation and annealing to generate the ensemble NV centres, and a concentration of NV⁻ centres as high as 10¹⁶ cm⁻³ in diamond was obtained. In addition, they found that sensitivity of the magnetometer using this method can reach as low as 5.22 µT/Hz currently.

This research can shed light on the development of an atomic magnetometer and a gyroscope on the basis of the ensemble spin of NV centres in diamond.

High concentration spin of NV⁻ in diamond is one of the advantages compared with that of the atomic vapor cells, because it can obtain a higher concentration. When increasing the spin concentration, the spin signal is easy to detect, and macro-atomic spin magnetometer become possible. This research is the first step for solid atomic magnetometers with high spin density and high sensitivity potentially with further optimization. It has a wide range of applications from fundamental physics tests, sensor applications and navigation to detection of NMR signals.

As has been pointed out, in this research, the authors mainly worked on fabricating NV⁻ centres with high concentration (10¹⁵-10¹⁶ cm⁻³) in diamond by using optimal irradiation and annealing processes, and they quantitatively defined the NV⁻ concentration, which is important for the design of higher concentration processes in the magnetometer and gyroscope. Until now, few groups can directly define the NV⁻ concentration. Besides, the authors optimized the microwave antenna parameters experimentally and explored the dependence between the splitting of the magnetic resonance and the magnetic fields, which dictated the minimum detectable magnetic field.

As manufacturing technology has developed, digital models from advanced measuring devices have been widely used in manufacturing sectors. To speed up the production cycle and reduce extra errors introduced in surface reconstruction processes, directly machining digital models in the polygonal stereolithographyformat has been considered as an effective approach in rapid digital manufacturing. In machining processes, Cutter Location (CL) data for numerical control (NC) machining is generated usually from an offset model. This model is created by offsetting each vertex of the original model along its vertex vector. However, this method has the drawback of overcut to the offset model. The purpose of this paper is to solve the overcut problem through an error compensation algorithm to the vertex offset model.

Based on the analysis of the vertex offset method and the offset model generated, the authors developed and implemented an error compensation method to correct the offset models and generated the accurate CL data for the subsequent machining process. This error compensation method is verified through three polygonal models and the tool paths generated were used for a real part machining.

Based on the analysis of the vertex offset method and the offset model generated, the authors developed an error compensation method to correct the offset models and generated the accurate CL data for the subsequent machining process. The developed error compensation algorithm can effectively solve the overcut drawback of the vertex offset method.

The error compensation method to the vertex offset model is used for generating the CL data with the using of a ball-end cutter.

On the study of CL data generation for a STL model, most of the current studies are focused on the determination of the offset vectors of the vertexes. The offset distance is usually fixed to the radius of the cutter used. Thus, the overcut problem to the offset model is inevitable and has not been much studied. The authors propose an effective approach to compensate the insufficient distance of the offset vertex and solve the overcut problem.

The directly tool paths generation from a STL model can reduce the error of surface reconstruction and speed up the machining progress.

The authors investigate the overcut problem occurred in vertex offset for CL data generation and present a new error compensation algorithm for generating the CL data that can effectively solve the overcut problem.