Density based semi-automatic labeling on multi-feature representations for ground truth generation: Application to handwritten character recognition

作者：

Highlights：

• Effective semi-automatic labeling method applying to ground truth generation for OCR.

• Alternative of parameters tuning for the proposed method.

• Evaluation of labeling performance of the proposed method on practical datasets.

摘要

•Effective semi-automatic labeling method applying to ground truth generation for OCR.•Alternative of parameters tuning for the proposed method.•Evaluation of labeling performance of the proposed method on practical datasets.

论文关键词：Semi-supervised learning,Active learning,Ground truth generation,Handwritten character recognition

论文评审过程：Received 1 September 2020, Revised 2 February 2021, Accepted 12 March 2021, Available online 15 March 2021, Version of Record 20 March 2021.

论文官网地址：https://doi.org/10.1016/j.knosys.2021.106953