Scalsol
diff --git a/‎.gitignore
-2 b/‎.gitignore
-2
diff --git a/‎INSTALL.md
+1-1 b/‎INSTALL.md
+1-1
diff --git a/‎README.md
+1-1 b/‎README.md
+1-1
diff --git a/‎datasets/ILSVRC2015/ImageSets/DET_train_30classes.txt
+53,639 b/‎datasets/ILSVRC2015/ImageSets/DET_train_30classes.txt
+53,639
diff --git a/‎datasets/ILSVRC2015/ImageSets/VID_train_15frames.txt
+57,834 b/‎datasets/ILSVRC2015/ImageSets/VID_train_15frames.txt
+57,834
diff --git a/‎datasets/ILSVRC2015/ImageSets/VID_train_every10frames.txt
+113,666 b/‎datasets/ILSVRC2015/ImageSets/VID_train_every10frames.txt
+113,666
@@ -32,6 +32,4 @@ dist/
 .DS_Store
 
 # project dirs
-/datasets
 /models
-/output
@@ -62,7 +62,7 @@ cd mega.pytorch
 # re-build it
 python setup.py build develop
 
-pip install 'pillow<=7.0.0'
+pip install 'pillow<7.0.0'
 
 unset INSTALL_DIR
 
 
@@ -6,7 +6,7 @@ By Yihong Chen, Yue Cao, Han Hu, Liwei Wang.
 
 ## Introduction
 
-This project is the codebase for implementing our MEGA algorithm for video object detection presented in our CVPR2020 paper.
+This project is the codebase based on [maskrcnn_benchmark](https://github.com/facebookresearch/maskrcnn-benchmark) for implementing our MEGA algorithm for video object detection presented in our CVPR2020 paper.
 
 The full paper is available at: