Search results

The paper aims to tackle the classification of Remote Sensing Images (RSIs), which presents a significant challenge for computer algorithms due to the inherent characteristics of clustered ground objects and noisy backgrounds. Recent research typically leverages larger volume models to achieve advanced performance. However, the operating environments of remote sensing commonly cannot provide unconstrained computational and storage resources. It requires lightweight algorithms with exceptional generalization capabilities.

This study introduces an efficient knowledge distillation (KD) method to build a lightweight yet precise convolutional neural network (CNN) classifier. This method also aims to substantially decrease the training time expenses commonly linked with traditional KD techniques. This approach entails extensive alterations to both the model training framework and the distillation process, each tailored to the unique characteristics of RSIs. In particular, this study establishes a robust ensemble teacher by independently training two CNN models using a customized, efficient training algorithm. Following this, this study modifies a KD loss function to mitigate the suppression of non-target category predictions, which are essential for capturing the inter- and intra-similarity of RSIs.

This study validated the student model, termed KD-enhanced network (KDE-Net), obtained through the KD process on three benchmark RSI data sets. The KDE-Net surpasses 42 other state-of-the-art methods in the literature published from 2020 to 2023. Compared to the top-ranked method’s performance on the challenging NWPU45 data set, KDE-Net demonstrated a noticeable 0.4% increase in overall accuracy with a significant 88% reduction in parameters. Meanwhile, this study’s reformed KD framework significantly enhances the knowledge transfer speed by at least three times.

This study illustrates that the logit-based KD technique can effectively develop lightweight CNN classifiers for RSI classification without substantial sacrifices in computation and storage costs. Compared to neural architecture search or other methods aiming to provide lightweight solutions, this study’s KDE-Net, based on the inherent characteristics of RSIs, is currently more efficient in constructing accurate yet lightweight classifiers for RSI classification.

When the emergency transition started in the spring of 2020 in the USA, teachers had to quickly switch from traditional in-person teaching to distance and remote teaching, regardless of their level of preparation. The distance and remote learning environments and contexts were different from traditional classrooms, which significantly changed the way teachers communicated and engaged with students in learning. The purpose of this study was to explore the workplace learning experience of K-12 educators during their work transition due to the COVID-19 pandemic in the USA.

In total, 30 qualitative, in-depth, semi-structured, one-on-one interviews were conducted with K-12 educators in Arkansas in the USA and synthesized their experiences.

This study identified four major themes in the workplace learning experiences of K-12 teachers during the COVID-19 pandemic: major challenges in workplace learning, including limited time, information overload, lack of relevance and customization and balancing priorities; challenges specific to different subgroups, such as age differences, prior experience and access to technology; strategies of workplace learning, notably collaborative learning, on-the-job training and professional development; and support for workplace learning, encompassing both formal support from schools and districts and informal support from family, friends and peers.

The paper provides original insights into K-12 teachers’ workplace learning during the COVID-19 pandemic by understanding their adaptation strategies. It fills a research gap by highlighting both the challenges and support mechanisms in educational transitions during a crisis.

This study makes an enquiry of the existing sustainable development goals (SDGs) knowledge of English language teacher trainers (ELTTs), who remain an overlooked stakeholder in education for sustainable development. Despite the literature on SDG integration into traditional teacher education curriculum, how massive open online courses (MOOCs) can help with capacity building of ELTTs is unclear. This study aims to further explore how the knowledge and capacities of ELTTs on SDGs could be promoted through MOOCs.

For the study’s action research approach, data of 28 ELTTs from higher education institutions in Ankara, Türkiye, were collected for six weeks using a survey, reflective reports and focus group interviews.

ELTTs with no previous training experience have a partially fragmented understanding of SDGs. MOOCs proved to be highly practical and cost-effective in building ELTTs’ capacity for building knowledge of and improving motivation to address SDGs in training programs. Further findings are drawn from ELTTs’ reflections on the lack of pedagogical content knowledge and collegial interactions in MOOCs.

There is little to no literature on ELTTs’ existing knowledge of SDGs and whether MOOCs can be used for their capacity building to achieve SDGs. The present findings may encourage higher education institutions to launch professional development programs that will equip ELTTs with the competencies to apply education for sustainable development in in-service English language teaching and teacher training programs.

A crucial issue in educational management refers to helping teachers reach their full potential and manage their talents. Although managing talent is advised as an essential resource for organizational transformation to maximize performance and to promote a school’s knowledge capital increase, Teachers’ talent management (TTM) is not an agreed-upon concept in the literature. The aim of this study is threefold: (1) to propose a theoretical and operational TTM definition, (2) to develop a test to measure TTM and (3) to discuss the value of TTM implementation in educational settings.

A psychometric validation design according to international standards for test development is implemented. A cross-sectional intentional sampling with 200 schools is used for obtaining reliability and validity evidence. We operationalized TTM using 9 factors (10 items per factor, 6-point Likert scale): attraction, selection, development, retention, succession, climate, culture, evaluation and knowledge management.

Results show adequate evidence of reliability and validity, focusing on items' content and latent structure. The combined use of structural equation models (SEM) and psychometric networks (PN) detects a six-factor model with a test reduced version including 61 items, getting 64.28% of the total explained variance and adequate reliability indices and factor loadings.

The Teachers' Talent Management Assessment Test (TTMAT) can contribute educational institutions to create a culture of excellence and deliver the best possible education to their students. Educational transformation will only happen if teachers are professionalized, trained, motivated and supported to inspire and to guide their learners to reach their objectives and well-being.

In Germany, various approaches have been taken to tackle the current teacher shortage in technical and vocational education and training (TVET). One attempt to remedy the shortage in Bavaria has been the introduction of an engineering education study programme at universities of applied sciences. Ideal candidates for this programme should have an interest in both engineering and social interaction. For effective recruitment, therefore, it is necessary to know applicants’ characteristics such as their vocational interests. In this study, the vocational interest profiles of students in TVET teacher training programmes were identified and their interest profiles and further characteristics were compared with those of other VET students at universities and universities of applied sciences.

An online questionnaire based on Holland’s interest theory and adapted from the Allgemeiner-Interessen-Struktur-Test-3 (interest structure test) was administered to 85 students in TVET teacher training programmes at universities and universities of applied sciences in Bavaria. Items regarding reasons for choosing a particular study programme, university location and other personal details were added.

The vocational interest profiles of students at universities and universities of applied sciences can be described as similar but weakly differentiated. Insights are provided by the characteristics of students such as the majority being first-time academics in the family. The reasons for choosing the degree programme and university location highlight the fact that a large proportion of students in engineering education would not have chosen a teaching-related degree programme if it had not been offered at the respective university of applied sciences.

Although the sample in this study was small and, therefore, limiting, it represented a high proportion of TVET teacher training students in Bavaria and a substantial proportion of first-year students in TVET teacher training programmes at universities and universities of applied sciences in Bavaria (section 2.2 and 3.1). Thus, the findings provide valuable insights into commonalities in interest profiles between engineering education students at universities of applied sciences and other TVET students at universities. With respect to the domain of the chosen vocational specialisation, differentiated profiles emerged that, for example, showed a stronger artistic orientation among students in construction technology/wood. For further analysis, the previous variable-centred orientation of the analysis can be supplemented by person-centred analyses (e.g. cluster analysis and latent variable mixture modelling, LVMM) (cf. Leon et al., 2021).

The findings in this study reveal the potential for attracting candidates to universities of applied sciences if they prefer to study in rather rural areas close to their hometowns. With the aim to educate prospective teachers for future work not only in metropolitan regions but in rural areas too, offering bachelor degree programmes in rural areas would seem promising. A regional option can boost the recruitment of new students and attract candidates that otherwise would be unable to pursue studies or a career as a teacher in vocational education. The results of this study and those of previous studies suggest that universities of applied sciences can cooperate with universities to help solve the teacher shortage problem.

Overall, it is apparent that the students' interests reached comparatively high values in all interest orientations and thus are only weakly differentiated. If undifferentiated profiles indicate low levels of career readiness, this significantly affects the recruitment of young people for the teaching profession. Assessing career orientation and promoting vocational interests should be prioritised during secondary school education. Vocational orientation measures are essential and should provide insight into typical activities of daily work life in different professions and thus pique and foster interests.

This study provides insight into how to respond to the teacher shortage in VET by identifying important characteristics of engineering education students using vocational interest profiling.