RIMES

该数据集名为Rimes，它被用于评估新型循环神经网络细胞设计在手写识别任务中的性能表现。此外，为了评估所提议的长短时记忆细胞替代方案的表现，该数据集还与IFN/ENIT数据集进行了比较。该数据集的主要任务是转录图像上的手写文本。

This dataset, named Rimes, is utilized to evaluate the performance of novel recurrent neural network cell designs for handwritten text recognition tasks. Furthermore, to assess the performance of the proposed long short-term memory (LSTM) cell alternatives, this dataset was compared against the IFN/ENIT dataset. The core task of this dataset is to transcribe handwritten text from images.