Summary of Namer: Non-autoregressive Modeling For Handwritten Mathematical Expression Recognition, by Chenyu Liu et al.
NAMER: Non-Autoregressive Modeling for Handwritten Mathematical Expression Recognition
by Chenyu Liu, Jia Pan, Jinshui Hu, Baocai Yin, Bing Yin, Mingjun Chen, Cong Liu, Jun Du, Qingfeng Liu
First submitted to arxiv on: 16 Jul 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Machine Learning (cs.LG)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary A novel bottom-up Non-AutoRegressive Modeling (NAMER) approach is proposed for Handwritten Mathematical Expression Recognition (HMER), which outperforms current state-of-the-art methods on ExpRate by 1.93%/2.35%/1.49%/0.62% and achieves significant speedups of 13.7x and 6.7x faster in decoding time and overall FPS. The NAMER framework consists of a Visual Aware Tokenizer (VAT) and a Parallel Graph Decoder (PGD), which tokenizes visible symbols and local relations at a coarse level, then refines all tokens and establishes connectivities in parallel, leveraging comprehensive visual and linguistic contexts. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary A new way to recognize handwritten math problems is developed. This method, called NAMER, is better than current methods at recognizing math expressions and does it much faster. The NAMER approach uses two parts: a Visual Aware Tokenizer that breaks down the problem into smaller pieces, and a Parallel Graph Decoder that puts those pieces together to create the final solution. |
Keywords
» Artificial intelligence » Autoregressive » Decoder » Tokenizer
