The Transformer-based encoder-decoder architecture has recently made
significant advances in recognizing handwritten mathematical expressions.
However, the transformer model still suffers from the lack of coverage problem,
making its expression recognition rate (ExpRate) inferior to its RNN
counterpart. Coverage information, which records the alignment information of
the past steps, has proven effective in the RNN models. In this paper, we
propose CoMER, a model that adopts the coverage information in the transformer
decoder. Specifically, we propose a novel Attention Refinement Module (ARM) to
refine the attention weights with past alignment information without hurting
its parallelism. Furthermore, we take coverage information to the extreme by
proposing self-coverage and cross-coverage, which utilize the past alignment
information from the current and previous layers. Experiments show that CoMER
improves the ExpRate by 0.61%/2.09%/1.59% compared to the current
state-of-the-art model, and reaches 59.33%/59.81%/62.97% on the CROHME
2014/2016/2019 test sets.

本文提出了使用注意力精细模块的覆盖信息 Transformer 模型（CoMER），用于识别手写数学表达式，实验显示 CoMER 相对于当前最先进的模型提高了 0.61%/2.09%/1.59% 的表达式识别率，并在 CROHME2014/2016/2019 测试集上达到了 59.33%/59.81%/62.97%。

CoMER: 基于 Transformer 的手写数学表达式识别覆盖建模

CoMER: Modeling Coverage for Transformer-based Handwritten Mathematical  Expression Recognition

In NMT, words are sometimes dropped from the source or generated repeatedly
in the translation. We explore novel strategies to address the coverage problem
that change only the attention transformation. Our approach allocates
fertilities to source words, used to bound the attention each word can receive.
We experiment with various sparse and constrained attention transformations and
propose a new one, constrained sparsemax, shown to be differentiable and
sparse. Empirical evaluation is provided in three languages pairs.

本文提出了一种新颖的方法来解决 NMT 中的 “coverage problem”，通过给源语言单词分配 “fertilities” 来限制每个单词可以接收的注意力，同时提出了一种新的稀疏的注意力变换方式 “sparsemax”，并在三种语言对上进行实证评估。