Merge pull request #5 from Haru0517/docs-patch

Update README.md

README.md

@@ -14,15 +14,20 @@ Make the executable file (`extract_nfeatures`) for [DeepSentiBank](https://githu
 You can get the executable file (`extract_nfeatures`) in source folder.
 
 ## Run in DeepSentiBank
-Download `caffe_sentibank_train_iter_250000` file from [this url](https://www.dropbox.com/s/lv3p67m21kr3mrg/caffe_sentibank_train_iter_250000?dl=1).
+Download `caffe_sentibank_train_iter_250000` file from [HERE](https://www.dropbox.com/s/lv3p67m21kr3mrg/caffe_sentibank_train_iter_250000?dl=1).
 
 Then, add `extract_nfeatures` and `caffe_sentibank_train_iter_250000` file to your [DeepSentiBank](https://github.com/generalmilk/DeepSentiBank) folder, and run `python sentiBank.py test_image.jpg`. 
 
 You can get output json file: `test_image.json`.
 
 ## Note
-(2020.01.24) Original `extract_nfeatures.cpp` in [DeepSentiBank](https://github.com/generalmilk/DeepSentiBank) contains a bug caused by [Caffe](https://github.com/BVLC/caffe) updates. (Look this [Issue](https://github.com/BVLC/caffe/issues/4107).)
+#### (2020.01.24) 
+Original `extract_nfeatures.cpp` in [DeepSentiBank](https://github.com/generalmilk/DeepSentiBank) contains a bug caused by [Caffe](https://github.com/BVLC/caffe) updates. (Look this [Issue](https://github.com/BVLC/caffe/issues/4107).)
 
 So we use a modified file (`source/extract_nfeatures.cpp`) instead of the original one.
 
+#### (2021.07.15) 
+The source of `caffe_sentibank_train_iter_250000` download link is [HERE](https://github.com/ColumbiaDVMM/ColumbiaImageSearch/blob/1326ee97ac1f032fdfbd5245c8356f59b254a9b5/cufacesearch/cufacesearch/featurizer/sbcmdline_img_featurizer.py#L18).
+
+