diff --git a/doc/doc_ch/recognition.md b/doc/doc_ch/recognition.md
index c2b61a28023e75a09c0cd2a273ec7451057c8240..80a8fb13f2d04961041211e83ef115842ff7054c 100644
--- a/doc/doc_ch/recognition.md
+++ b/doc/doc_ch/recognition.md
@@ -48,14 +48,15 @@ mklink /d <path/to/paddle_ocr>/train_data/dataset <path/to/dataset>
 ```
 " 图像文件名                 图像标注信息 "
 
-train_data/train/word_001.jpg   简单可依赖
-train_data/train/word_002.jpg   用科技让复杂的世界更简单
+train_data/rec/train/word_001.jpg   简单可依赖
+train_data/rec/train/word_002.jpg   用科技让复杂的世界更简单
 ...
 ```
 
 最终训练集应有如下文件结构：
 ```
 |-train_data
+  |-rec
     |- rec_gt_train.txt
     |- train
         |- word_001.png
@@ -70,6 +71,7 @@ train_data/train/word_002.jpg   用科技让复杂的世界更简单
 
 ```
 |-train_data
+  |-rec
     |- rec_gt_test.txt
     |- test
         |- word_001.jpg
diff --git a/doc/doc_en/recognition_en.md b/doc/doc_en/recognition_en.md
index f6c4c1059acc767253ee3b0477eb16924651ce3e..483df5d7ef58337b0b4e6a2453aefb6a2ca80f24 100644
--- a/doc/doc_en/recognition_en.md
+++ b/doc/doc_en/recognition_en.md
@@ -49,8 +49,8 @@ It is recommended to put the training images in the same folder, and use a txt f
 ```
 " Image file name           Image annotation "
 
-train_data/train/word_001.jpg   简单可依赖
-train_data/train/word_002.jpg   用科技让复杂的世界更简单
+train_data/rec/train/word_001.jpg   简单可依赖
+train_data/rec/train/word_002.jpg   用科技让复杂的世界更简单
 ...
 ```
 
@@ -58,6 +58,7 @@ The final training set should have the following file structure:
 
 ```
 |-train_data
+  |-rec
     |- rec_gt_train.txt
     |- train
         |- word_001.png
@@ -72,6 +73,7 @@ Similar to the training set, the test set also needs to be provided a folder con
 
 ```
 |-train_data
+  |-rec
     |-ic15_data
         |- rec_gt_test.txt
         |- test