我自己的结果类似10,10,100,10,100,50,10,50, happy如何使用字典 how to use the lexicon in my result?

[ustczhouyu]

我的端到端识别结果中每个txt文件的每一行类似于：10,10,100,10,100,50,10,50, happy
分别对应坐标和识别的单词。请问这样的结果怎么使用S，W，G的lexicon?

Each line of every txt file in my end-to-end recognition result is similar to: 10,10,100,10,100,50,10,50, happy
Correspond to coordinates and recognized words respectively. How do I use S, W, G lexicon for such results?

@MhLiao @lcsama