delveintodetail

有， 效果很好，在平安财产保险公司，大概有40-50个单证（身份证，行驶证，驾驶证，银行卡，出生证明，死亡证明，医疗发票，营业执照.......）都是用Master做的， 调用量年均2.4亿次。

哈哈， cpu估计训练起来基本速度肯定不行， 是一定要GPU的。。。

我们报告的这个是表格序列预测的准确性， 这个指标比较严格，要求序列全对才是对，否则纠错，有些结构的序列长度会是300-500个token， 一个都不错不是很容易的事情，78%的准确率其实是不低的，我大致记得，结构序列准确性只有60%多一点， 最后的ted都有可能达到96%，Teds是一个很松的指标，跟它的计算方式有关。即使结构序列错了一个token， 可能这张图的Teds也可以达到97%以上。。

According to my guess, the performance of this implementation should be 85% on IIIT-5K.

> @delveintodetail have you trained. > the developer didnot reply clearly in the matter of training. > whether he crops the icdar words, or what.... It is not because of...

对于空表格，只需要分类token， 不回归框的位置。。。用mask

可以关注一下好未来的表格识别比赛的数据， 和一些人的相关方案。。。

In the released code, we did not put the speedup module in the repo. You can refer to our MASTER paper for the speedup module. I expect the end2end inference...

In the self-regressed decoder, there are many repeated operations. In the master paper, we use a memory-cached mechanism to speedup the inference. The speedup is extremely effecient for the long...

In the competition, we used our own internal tool (FastOCR) to implement our algorithm. In the mmocr framework, we did not implement it. If you fully understand the master paper,...