diff --git a/cvpr_paper_download.py b/cvpr_paper_download.py
new file mode 100644
index 0000000..940ba8a
--- /dev/null
+++ b/cvpr_paper_download.py
@@ -0,0 +1,86 @@
+# -*- coding: utf-8 -*-
+"""
+Created on Sat Nov 17 17:11:28 2018
+
+@author: Zhiming Zhang
+
+mail: i@dropout.cn
+
+"""
+
+import os
+import re
+import requests
+from bs4 import BeautifulSoup
+
+
+log=r'E:\Dataset\book_test'
+save_dir=r'E:\Dataset\book_test'
+#save_dir is the default download dir
+
+def get_url_name(content,year_num):
+    year=str(year_num)
+    book_url_list=[]
+    url='http://openaccess.thecvf.com/{}{}.py'.format(content,year)
+    r=requests.get(url)
+    html=r.text
+    soup=BeautifulSoup(html,'html.parser')
+    dds=soup.find_all('dd')
+    for i,dd in enumerate(dds):
+        if i%2==1:
+            href=dd.find('a').get('href')
+            book_url=r'http://openaccess.thecvf.com/'+href
+            book_url_list.append(book_url)
+    return book_url_list
+
+def download(content,year_num,book_url_list,save_dir):
+    year=str(year_num)
+    root_dir=os.path.join('{}_{}'.format(content,year))
+    print('downloading  the {} content papar of {} year'.format(content,year))
+    if not os.path.exists(root_dir):
+        os.makedirs(root_dir) 
+    total_num=len(book_url_list)
+    print('{} content of {} year total is {} '.format(content,year,total_num))
+    
+    key_interrupt=0
+    for i,url in enumerate(book_url_list):
+        name=url.split('/')[-1]
+        ab_name=os.path.join(root_dir,name)
+        if os.path.exists(ab_name):
+            print('{} has already existed'.format(name))
+            return
+        try:
+            r=requests.get(url,stream=True)
+            with open(ab_name,'wb') as f:
+                print('Now writing the {} file ：{}'.format(i+1,name))
+                for chunk in r.iter_content(chunk_size=4096):
+                    if chunk:
+                        f.write(chunk)
+                        f.flush
+            r.close()
+        except KeyboardInterrupt:
+            key_interrupt=1
+            if os.path.exists(ab_name):
+                os.remove(ab_name)
+            return key_interrupt
+        except Exception:
+            key_interrupt=1
+            return key_interrupt
+    print('{} content of {} year has downloaded successfully'.format(content,year))
+
+
+def run():
+    key_interrupt=0
+    year_content_dict={'CVPR':[year for year in range(2015,2017)],'ICCV':[2013,2015,2017]}
+    for content,year_list in year_content_dict.items():
+        for year in year_list:
+            book_url_list=get_url_name(content,year)
+            key_interrupt=download(content,year,book_url_list,save_dir)
+            if key_interrupt:
+                print('keyboard interrupt')
+                return
+        
+if __name__=='__main__':
+    run()
+
+