0930 ECL車牌辨識

This license plate recognition is one of the example applications of monotonic improvement technology. Traditional supervised learning requires the collection of new data and human resources to mark these new data, and then the new data can be used to train the model; however, this monotonic improvement technology uses semi-supervised learning which can be used to train the model without spending extra human resources to mark the newly collected data to achieve the ability of automatic learning and optimization of the model.

 

本車牌辨識為單調提升技術其中一項範例應用，傳統監督式學習需要將圖資蒐集後花費人力資源來作圖資標記，才能將新圖資用於訓練模型；然而，本單調提升技術採半監督式學習，不需花費額外的人力資源標記新蒐集的圖資，即可用於訓練模型，來達到模型自動學習優化的能力。