Search results

The defect detection problem of color-patterned fabric is still a huge challenge due to the lack of manual defect labeling samples. Recently, many fabric defect detection algorithms based on feature engineering and deep learning have been proposed, but these methods have overdetection or miss-detection problems because they cannot adapt to the complex patterns of color-patterned fabrics. The purpose of this paper is to propose a defect detection framework based on unsupervised adversarial learning for image reconstruction to solve the above problems.

The proposed framework consists of three parts: a generator, a discriminator and an image postprocessing module. The generator is able to extract the features of the image and then reconstruct the image. The discriminator can supervise the generator to repair defects in the samples to improve the quality of image reconstruction. The multidifference image postprocessing module is used to obtain the final detection results of color-patterned fabric defects.

The proposed framework is compared with state-of-the-art methods on the public dataset YDFID-1(Yarn-Dyed Fabric Image Dataset-version1). The proposed framework is also validated on several classes in the MvTec AD dataset. The experimental results of various patterns/classes on YDFID-1 and MvTecAD demonstrate the effectiveness and superiority of this method in fabric defect detection.

It provides an automatic defect detection solution that is convenient for engineering applications for the inspection process of the color-patterned fabric manufacturing industry. A public dataset is provided for academia.

This paper aims to develop a model that reflects the current transformer (CT) core materials nonlinearity. The model enables simulation and analysis of the CT excitation current that includes the inductive magnetizing current and the resistive excitation current.

A nonlinear CT model is established with the magnetizing current as the solution variable. This model presents the form of a nonlinear differential equation and can be solved discretely using the Runge–Kutta method.

By simulating variations in the excitation current for different primary currents, loads and core materials, the results demonstrate that enhancing the permeability of the B–H curve leads to a more significant improvement in the CT ratio error at low primary currents.

The proposed model has three obvious advantages over the previous models with the secondary current as the solution variable: (1) The differential equation is simpler and easier to solve. Previous models contain the time differential terms of the secondary current and excitation flux or the integral term of the flux, making the iterative solution complicated. The proposed model only contains the time differential of the magnetizing current. (2) The accuracy of the excitation current obtained by the proposed model is higher. Previous models calculate the excitation current by subtracting the secondary current from the converted primary current. Because these two currents are much greater than the excitation current, the error of calculating the small excitation current by subtracting two large numbers is greatly enlarged. (3) The proposed model can calculate the distorted waveform of the excitation current and error for any form of time-domain primary current, while previous models can only obtain the effective value.

This study examines the effects of scent arousal on consumers' time perception in retail service environments and further explores how the effect is moderated by consumer-perceived stress.

A laboratory experiment (Study 1) and a field experiment (Study 2) were conducted to examine the relationship between scent arousal and time perception and the mediating effect between scent arousal and consumers' store evaluations. Another laboratory experiment (Study 3) was conducted to explore how consumers' stress modifies the scent arousal effect.

Consumers in a low-arousal scent condition perceived a shorter duration of time than those in a high-arousal scent condition. This finding was verified in a field experiment, whereas scent arousal affects consumers' store evaluations through the mediating effects of time perception. However, the impact of scent arousal on time perception was attenuated in high-stress conditions.

Time duration perception is an important indicator in the retail service marketing process. Evidence shows that underestimating time duration in the shopping process represents positive responses. This study extends prior research by examining how scent arousal influences time perception and how consumers' stress moderates scent arousal’s effect.

This paper aims to automatically obtain the implant parameter from the CBCT images to improve the outcome of implant planning.

This paper proposes automatic simulated dental implant positioning on CBCT images, which can significantly improve the efficiency of implant planning. The authors introduce the fusion point calculation method for the missing tooth's long axis and root axis based on the dental arch line used to obtain the optimal fusion position. In addition, the authors proposed a semi-interactive visualization method of implant parameters that be automatically simulated by the authors' method. If the plan does not meet the doctor's requirements, the final implant plan can be fine-tuned to achieve the optimal effect.

A series of experimental results show that the method proposed in this paper greatly improves the feasibility and accuracy of the implant planning scheme, and the visualization method of planting parameters improves the planning efficiency and the friendliness of system use.

The proposed method can be applied to dental implant planning software to improve the communication efficiency between doctors, patients and technicians.

Music sentiment analysis helps to promote the diversification of music information retrieval methods. Traditional music emotion classification tasks suffer from high manual workload and low classification accuracy caused by difficulty in feature extraction and inaccurate manual determination of hyperparameter. In this paper, the authors propose an optimized convolution neural network-random forest (CNN-RF) model for music sentiment classification which is capable of optimizing the manually selected hyperparameters to improve the accuracy of music sentiment classification and reduce labor costs and human classification errors.

A CNN-RF music sentiment classification model is designed based on quantum particle swarm optimization (QPSO). First, the audio data are transformed into a Mel spectrogram, and feature extraction is conducted by a CNN. Second, the music features extracted are processed by RF algorithm to complete a preliminary emotion classification. Finally, to select the suitable hyperparameters for a CNN, the QPSO algorithm is adopted to extract the best hyperparameters and obtain the final classification results.

The model has gone through experimental validations and achieved a classification accuracy of 97 per cent for different sentiment categories with shortened training time. The proposed method with QPSO achieved 1.2 and 1.6 per cent higher accuracy than that with particle swarm optimization and genetic algorithm, respectively. The proposed model had great potential for music sentiment classification.

The dual contribution of this work comprises the proposed model which integrated two deep learning models and the introduction of a QPSO into model optimization. With these two innovations, the efficiency and accuracy of music emotion recognition and classification have been significantly improved.

The coordinated development of the urbanization and construction industry is crucial for the sustainable development of cities. However, the coupling relationship and coordination mechanism between them remain unclear. To bridge this gap, this study attempts to explore the level of coupling coordination between new urbanization and construction industry development and investigate the critical driving factors influencing their coupling coordination degree.

By referring to the existing literature, two index systems were established to evaluate the development level of the new urbanization and construction industry. The spatiotemporal characteristics of the coupled coordinated development of the new urbanization and construction industry in China from 2014 to 2020 were investigated using the coupling coordination model. The Markov chain and geographic detector were adopted to understand the transition probability and driving factors of the coupling coordination degree.

The results indicate that the coupling degree of China's new urbanization and construction industry is high, and the two systems exhibit obvious interaction phenomena. However, the construction industry in most provinces lags behind the new urbanization. A positive interactive relationship and coordination mechanism has not been established between the two systems. Furthermore, the  coupling contribution degree of the driving factors from high to low is as follows: market size > labor resource concentration > government investment ability > economic development level > industrial structure > production efficiency > technology level. Accordingly, a driving mechanism including market, policy, economic, and production technology drivers was developed.

This study contributes to the existing body of knowledge by providing a set of scientific analysis methods to address the deficiency of coordination mechanism research on new urbanization and the construction industry. The results also provide a theoretical basis for decision makers to develop differentiated sustainable development policies.

The ability of construction firms to become more environmentally conscious and socially responsible for their business activities has been touted as the key driver for improved individual firms' competitiveness. This study explores the key dimensions of corporate social responsibility (CSR) practices and their impact factors among construction firms.

Through the institutional, stakeholders and self-determination theories, this study proposed a conceptual framework of CSR implementation. For its validation, data were collected from 90 top-tier construction firms using an online survey and analyzed via a two-pronged factor analysis method.

The empirical results demonstrate that the CSR practices of construction firms include eight key dimensions, e.g. shareholders' interests, government commitment and CSR institutional arrangement. The three key groups of impact factor of CSR implementation are (1) identified factors (i.e. contractors' perceived importance of CSR practices); (2) external institutional factors (i.e. coercive and normative factors and mimetic factors); and (3) intrinsic factors (i.e. strategic business direction, resource and capability and organizational culture).

The research findings inform the practitioners about how to enact, manage and improve firms' socially responsible goals so as to fulfill their key stakeholders' requirements and expectations and thus enhance their legitimacy in construction businesses.

This study contributes to CSR knowledge by identifying and empirically demonstrating valid measurements of the key dimensions of practices and impact factors toward CSR implementation by construction firms.