Add explanation for using denseflow to extract both rgb and flow (ope…

…n-mmlab#17)

docs/data_preparation.md

@@ -41,14 +41,15 @@ Note that the videos should be arranged in either
 
 ### Extract frames
 
-To extract frames and optical flow, [dense_flow](https://github.com/open-mmlab/denseflow) is needed.
-You can use the following command to extract frames.
+To extract both frames and optical flow, you can use the tool [denseflow](https://github.com/open-mmlab/denseflow) we wrote.
+Since different frame extraction tools produce different number of frames,
+it is beneficial to use the same tool to do both frame extraction and the flow computation, to avoid mismatching of frame counts.
 
 ```shell
 python build_rawframes.py ${SRC_FOLDER} ${OUT_FOLDER} [--task ${TASK}] [--level ${LEVEL}] \
     [--num-worker ${NUM_WORKER}] [--flow-type ${FLOW_TYPE}] [--out-format ${OUT_FORMAT}] \
     [--ext ${EXT}] [--new-width ${NEW_WIDTH}] [--new-height ${NEW_HEIGHT}] [--new-short ${NEW_SHORT}]
-    [--resume]
+    [--resume] [--use-opencv]
 ```
 
 - `SRC_FOLDER`: Folder of the original video.
@@ -63,6 +64,7 @@ python build_rawframes.py ${SRC_FOLDER} ${OUT_FOLDER} [--task ${TASK}] [--level
 - `NEW_HEIGHT`: Resized image height of output.
 - `NEW_SHORT`: Resized image short side length keeping ratio.
 - `--resume`: Whether to resume optical flow extraction instead of overwriting.
+- `--use-opencv`: Whether to use OpenCV to extract rgb frames.
 
 The recommended practice is
 

tools/data/build_rawframes.py

@@ -139,10 +139,6 @@ def parse_args():
         default='avi',
         choices=['avi', 'mp4', 'webm'],
         help='video file extensions')
-    parser.add_argument(
-        '--use-opencv',
-        action='store_true',
-        help='Whether to use opencv to extract rgb frames')
     parser.add_argument(
         '--new-width', type=int, default=0, help='resize image width')
     parser.add_argument(
@@ -158,6 +154,10 @@ def parse_args():
         action='store_true',
         default=False,
         help='resume optical flow extraction instead of overwriting')
+    parser.add_argument(
+        '--use-opencv',
+        action='store_true',
+        help='Whether to use opencv to extract rgb frames')
     args = parser.parse_args()
 
     return args

tools/data/kinetics400/preparing_kinetics400.md

@@ -39,7 +39,7 @@ bash rename_classnames.sh
 
 This part is **optional** if you only want to use the video loader.
 
-Before extracting, please refer to [install.md](/docs/install.md) for installing [dense_flow](https://github.com/open-mmlab/denseflow).
+Before extracting, please refer to [install.md](/docs/install.md) for installing [denseflow](https://github.com/open-mmlab/denseflow).
 
 If you have plenty of SSD space, then we recommend extracting frames there for better I/O performance. And you can run the following script to soft link the extracted frames.
 

tools/data/mit/preparing_mit.md

@@ -15,7 +15,7 @@ bash download_data.sh
 
 This part is **optional** if you only want to use the video loader.
 
-Before extracting, please refer to [install.md](/docs/install.md) for installing [dense_flow](https://github.com/open-mmlab/denseflow).
+Before extracting, please refer to [install.md](/docs/install.md) for installing [denseflow](https://github.com/open-mmlab/denseflow).
 
 If you have plenty of SSD space, then we recommend extracting frames there for better I/O performance. And you can run the following script to soft link the extracted frames.
 

tools/data/mmit/preparing_mmit.md

@@ -15,7 +15,7 @@ bash download_data.sh
 
 This part is **optional** if you only want to use the video loader.
 
-Before extracting, please refer to [install.md](/docs/install.md) for installing [dense_flow](https://github.com/open-mmlab/denseflow).
+Before extracting, please refer to [install.md](/docs/install.md) for installing [denseflow](https://github.com/open-mmlab/denseflow).
 
 First, you can run the following script to soft link SSD.
 

tools/data/sthv1/preparing_sthv1.md

@@ -21,7 +21,7 @@ cd $MMACTION2/tools/data/sthv1/
 
 This part is **optional** if you only want to use the video loader.
 
-Before extracting, please refer to [install.md](/docs/install.md) for installing [dense_flow](https://github.com/open-mmlab/denseflow).
+Before extracting, please refer to [install.md](/docs/install.md) for installing [denseflow](https://github.com/open-mmlab/denseflow).
 
 If you have plenty of SSD space, then we recommend extracting frames there for better I/O performance.
 

tools/data/sthv2/preparing_sthv2.md

@@ -20,7 +20,7 @@ cat 20bn-something-something-v2-?? | tar zx
 
 This part is **optional** if you only want to use the video loader.
 
-Before extracting, please refer to [install.md](/docs/install.md) for installing [dense_flow](https://github.com/open-mmlab/denseflow).
+Before extracting, please refer to [install.md](/docs/install.md) for installing [denseflow](https://github.com/open-mmlab/denseflow).
 
 If you have plenty of SSD space, then we recommend extracting frames there for better I/O performance.
 

tools/data/thumos14/preparing_thumos14.md

@@ -25,7 +25,7 @@ bash download_videos.sh
 
 This part is **optional** if you only want to use the video loader.
 
-Before extracting, please refer to [install.md](/docs/install.md) for installing [dense_flow](https://github.com/open-mmlab/denseflow).
+Before extracting, please refer to [install.md](/docs/install.md) for installing [denseflow](https://github.com/open-mmlab/denseflow).
 
 If you have plenty of SSD space, then we recommend extracting frames there for better I/O performance.
 

tools/data/ucf101/preparing_ucf101.md

@@ -23,7 +23,7 @@ bash download_videos.sh
 
 This part is **optional** if you only want to use the video loader.
 
-Before extracting, please refer to [install.md](/docs/install.md) for installing [dense_flow](https://github.com/open-mmlab/denseflow).
+Before extracting, please refer to [install.md](/docs/install.md) for installing [denseflow](https://github.com/open-mmlab/denseflow).
 
 If you have plenty of SSD space, then we recommend extracting frames there for better I/O performance. The extracted frames (RGB + Flow) will take up about 100GB.