M²E: Multi-line Mathematical Formula Dataset

This dataset contains 99,956 multi-line mathematical expression images and their annotations. All images are taken from real-world scenes using mobile phones, and are multi-line mathematical formulas captured from math test papers and exercise books. Validation and test sets are specially divided to prevent overfitting during training. It can be used for mathematical formula recognition tasks. Related papers are "Read Ten Lines at One Glance: Line-Aware Semi-Autoregressive Transformer for Multi-Line Han".