English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Nowadays, the rapid development of photovoltaic(PV) power stations requiresincreasingly reliable maintenance and fault diagnosis of PV modules in thefield. Due to the effectiveness, convolutional neural network (CNN) has beenwidely used in the existing automatic defect detection of PV cells. However,the parameters of these CNN-based models are very large, which requirestringent hardware resources and it is difficult to be applied in actualindustrial projects. To solve these problems, we propose a novel lightweighthigh-performance model for automatic defect detection of PV cells inelectroluminescence(EL) images based on neural architecture search andknowledge distillation. To auto-design an effective lightweight model, weintroduce neural architecture search to the field of PV cell defectclassification for the first time. Since the defect can be any size, we designa proper search structure of network to better exploit the multi-scalecharacteristic. To improve the overall performance of the searched lightweightmodel, we further transfer the knowledge learned by the existing pre-trainedlarge-scale model based on knowledge distillation. Different kinds of knowledgeare exploited and transferred, including attention information, featureinformation, logit information and task-oriented information. Experiments havedemonstrated that the proposed model achieves the state-of-the-art performanceon the public PV cell dataset of EL images under online data augmentation withaccuracy of 91.74% and the parameters of 1.85M. The proposed lightweighthigh-performance model can be easily deployed to the end devices of the actualindustrial projects and retain the accuracy.

A lightweight network for photovoltaic cell defect detection in  electroluminescence images based on neural architecture search and knowledge  distillation