diff --git a/assignment-2/submission/16307130040/README.md b/assignment-2/submission/16307130040/README.md
index 74274e457388b56b4227c995ac2586137ddf1d19..75a75c96dba24c1fe572704318d3646659659437 100644
--- a/assignment-2/submission/16307130040/README.md
+++ b/assignment-2/submission/16307130040/README.md
@@ -4,7 +4,7 @@
 
 在完成了对mini_batch函数的替换后，模型得以顺利进行：
 
-![](.\img\Figure_1.png)
+![](./img/Figure_1.png)
 
 ```shell
 [0] Accuracy: 0.9453
@@ -43,7 +43,7 @@ def mini_batch(dataset, batch_size=128, numpy=True):
 
 matmul：
 
-#### ![](.\img\matmul.jpg)
+#### ![](./img/matmul.jpg)
 
 设输出y为l维的向量。
 
@@ -116,7 +116,7 @@ dyi/dxi=1/xi,dL/dxi=(dL/dyi)·1/xi
 
 4，softmax函数
 
-![](.\img\softmax.jpg)
+![](./img/softmax.jpg)
 
 设x和y均为l维的向量。