Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2603.07929 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917324276105216 |
|---|---|
| author | Le, Anh Duy Pham, Van Linh Ly, Vinh Loi Nguyen, Nam Quan Nguyen, Huu Thang Tran, Tuan Anh |
| author_facet | Le, Anh Duy Pham, Van Linh Ly, Vinh Loi Nguyen, Nam Quan Nguyen, Huu Thang Tran, Tuan Anh |
| contents | One of the crucial challenges taken in document analysis is mathematical expression recognition. Unlike text recognition which only focuses on one-dimensional structure images, mathematical expression recognition is a much more complicated problem because of its two-dimensional structure and different symbol size. In this paper, we propose using a Hybrid Vision Transformer (HVT) with 2D positional encoding as the encoder to extract the complex relationship between symbols from the image. A coverage attention decoder is used to better track attention's history to handle the under-parsing and over-parsing problems. We also showed the benefit of using the [CLS] token of ViT as the initial embedding of the decoder. Experiments performed on the IM2LATEX-100K dataset have shown the effectiveness of our method by achieving a BLEU score of 89.94 and outperforming current state-of-the-art methods. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_07929 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | A Hybrid Vision Transformer Approach for Mathematical Expression Recognition Le, Anh Duy Pham, Van Linh Ly, Vinh Loi Nguyen, Nam Quan Nguyen, Huu Thang Tran, Tuan Anh Computer Vision and Pattern Recognition One of the crucial challenges taken in document analysis is mathematical expression recognition. Unlike text recognition which only focuses on one-dimensional structure images, mathematical expression recognition is a much more complicated problem because of its two-dimensional structure and different symbol size. In this paper, we propose using a Hybrid Vision Transformer (HVT) with 2D positional encoding as the encoder to extract the complex relationship between symbols from the image. A coverage attention decoder is used to better track attention's history to handle the under-parsing and over-parsing problems. We also showed the benefit of using the [CLS] token of ViT as the initial embedding of the decoder. Experiments performed on the IM2LATEX-100K dataset have shown the effectiveness of our method by achieving a BLEU score of 89.94 and outperforming current state-of-the-art methods. |
| title | A Hybrid Vision Transformer Approach for Mathematical Expression Recognition |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2603.07929 |