Citation: Nam Van Hai Phan, Minh-Khoa Nguyen, Trung Thanh Nguyen, Trung Thanh Pham, Phuong-Nam Tran, and Duc Ngoc Minh Dang, “Mask CoMER: Enhancing Handwritten Mathematical Expression Recognition with Masked Language Pretraining and Regularization”, ICDAR - rank A, Wuhan, Hubei, China.
We are excited to share our latest research, “Mask CoMER: Enhancing Handwritten Mathematical Expression Recognition with Masked Language Pretraining and Regularization,” published by our team at AiTA Lab, FPT University, in collaboration with Kyung Hee University. This work introduces Mask CoMER, a novel two-stage model that significantly improves the recognition of handwritten mathematical expressions. By combining masked language model pretraining with stochastic depth regularization, Mask CoMER achieves state-of-the-art performance on the CROHME 2014, 2016, and 2019 datasets, with Expression Rates of 64.56%, 63.03%, and 65.22%, respectively. This advancement enhances applications in digitized education, document processing, and assistive technologies. Explore the code on GitHub and stay tuned for more updates from our lab!
,%20Wuhan,%20Hubei,%20China.){kind=link}