Unable to use extract flow

[niveditarahurkar]

I am trying to run -
scripts/extract_optical_flow.sh /temporal-segment-networks/src_folder/ /temporal-segment-networks/out_folder/ 1
I have taken just 5 sample videos from UCF101 in my src_folder:
v_SkateBoarding_g25_c01 v_SkateBoarding_g25_c02 v_SkateBoarding_g25_c03 v_SkateBoarding_g25_c04 v_SkateBoarding_g25_c05

I get the output:

0 v_SkateBoarding_g25_c05 done
4 v_SkateBoarding_g25_c01 done
1 v_SkateBoarding_g25_c02 done
3 v_SkateBoarding_g25_c03 done
2 v_SkateBoarding_g25_c04 done

But my out_folder --> out_folder/v_SkateBoarding_g25_c01# is empty.

My parameters look right:
('flow_type: ', 'tvl1') ('reading videos from folder: ', '/temporal-segment-networks/src_folder') ('selected extension of videos:', 'avi') ('total number of videos found: ', 5)

Not able to figure out what's going wrong. I reinstalled dense-flow as I thought some issue with build/extract_gpu but didn't help. Any ideas what could be the problem?

[raozhongyu]

I have met the same problem , did you solve it ? thank you

[a-whitej]

Met same problem. How to fix. On windows,it can be fixed by making copy opencv ffmpeg dll to current directory. but on linux, this kind of so file can not be found.

[a-whitej]

./extract_cpu -f=test.avi -x=tmp/flow_x -y=tmp/flow_y -i=tmp/image -b=20 -t=1 -d=0 -s=1 -o=dir
2019-08-23 15:58:13,896 FATAL [default] Check failed: [video_stream.isOpened()] Cannot open video stream "test.avi" for optical flow extraction.
2019-08-23 15:58:13,896 WARN [default] Aborting application. Reason: Fatal log at [/home/lisyi/dev/dense_flow/src/dense_flow.cpp:13]
[1] 30272 abort (core dumped) ./extract_cpu -f=test.avi -x=tmp/flow_x -y=tmp/flow_y -i=tmp/image -b=20 -t=1

[niveditarahurkar]

@raozhongyu and @a-whitej Did you try with the docker container that they have provided? You should not see any issues if you run the scripts from inside the container.

[a-whitej]

the problem is that opecv is not be installed correctly. during compiling opencv on problem i manually run compile such as
/usr/bin/c++ -isystem /usr/local/include/opencv -isystem /usr/local/include -I/home/my/dev/opencv-3.1.0/modules/cudaarithm/include -I/home/my/dev/opencv-3.1.0/modules/cudaoptflow/include -I/home/my/dev/opencv-3.1.0/modules/cudacodec/include -I/usr/lib/x86_64-linux-gnu/libzip/include -I/home/my/dev/dense_flow/include -I/home/my/dev/dense_flow/include/easylogging++/src -I/home/my/dev/dense_flow/include/CLUE/include -std=c++11 -fPIC -o CMakeFiles/denseflow.dir/src/dense_flow_gpu.cpp.o -c /home/my/dev/dense_flow/src/dense_flow_gpu.cpp
/usr/bin/c++ -isystem /usr/local/include/opencv -isystem /usr/local/include -I/home/my/dev/opencv-3.1.0/modules/cudaarithm/include -I/home/my/dev/opencv-3.1.0/modules/cudaoptflow/include -I/home/my/dev/opencv-3.1.0/modules/cudacodec/include -I/usr/lib/x86_64-linux-gnu/libzip/include -I/home/my/dev/dense_flow/include -I/home/my/dev/dense_flow/include/easylogging++/src -I/home/my/dev/dense_flow/include/CLUE/include -std=c++11 -fPIC -o CMakeFiles/denseflow.dir/src/dense_warp_flow_gpu.cpp.o -c /home/my/dev/dense_flow/src/dense_warp_flow_gpu.cpp

/usr/bin/c++ -Dpydenseflow_EXPORTS -isystem /usr/local/include/opencv -isystem /usr/local/include -I/usr/local/include/python3.5m -I/usr/lib/x86_64-linux-gnu/libzip/include -I/home/lisyi/dev/dense_flow/include -I/home/lisyi/dev/dense_flow/include/easylogging++/src -I/home/lisyi/dev/dense_flow/include/CLUE/include -std=c++11 -fPIC -fPIC -o CMakeFiles/pydenseflow.dir/src/py_denseflow.cpp.o -c /home/lisyi/dev/dense_flow/src/py_denseflow.cpp

then problem sovled.

[shijubushiju]

@niveditarahurkar I used Docker, but the following error occurred while running bash build_all.sh inside: can you help me fix it?
CMake Error: The following variables are used in this project, but they are set to NOTFOUND.
Please set them or make sure they are set and tested correctly in the CMake files:
LIBZIP_INCLUDE_DIR_ZIP
used as include directory in directory /Code/temporal-segment-networks/lib/dense_flow
used as include directory in directory /Code/temporal-segment-networks/lib/dense_flow
used as include directory in directory /Code/temporal-segment-networks/lib/dense_flow
used as include directory in directory /Code/temporal-segment-networks/lib/dense_flow
used as include directory in directory /Code/temporal-segment-networks/lib/dense_flow
used as include directory in directory /Code/temporal-segment-networks/lib/dense_flow
used as include directory in directory /Code/temporal-segment-networks/lib/dense_flow
LIBZIP_INCLUDE_DIR_ZIPCONF
used as include directory in directory /Code/temporal-segment-networks/lib/dense_flow
used as include directory in directory /Code/temporal-segment-networks/lib/dense_flow
used as include directory in directory /Code/temporal-segment-networks/lib/dense_flow
used as include directory in directory /Code/temporal-segment-networks/lib/dense_flow
used as include directory in directory /Code/temporal-segment-networks/lib/dense_flow
used as include directory in directory /Code/temporal-segment-networks/lib/dense_flow
used as include directory in directory /Code/temporal-segment-networks/lib/dense_flow
LIBZIP_LIBRARY
linked by target "denseflow" in directory /Code/temporal-segment-networks/lib/dense_flow
linked by target "extract_cpu" in directory /Code/temporal-segment-networks/lib/dense_flow
linked by target "extract_warp_gpu" in directory /Code/temporal-segment-networks/lib/dense_flow
linked by target "extract_gpu" in directory /Code/temporal-segment-networks/lib/dense_flow

[WangShuu]

@raozhongyu and @a-whitej Did you try with the docker container that they have provided? You should not see any issues if you run the scripts from inside the container.

我也是用docker来运行的，最后生成的同样是空文件夹，跟你的问题一样，你现在解决这个问题了嘛？

[tz0403]

@raozhongyu and @a-whitej Did you try with the docker container that they have provided? You should not see any issues if you run the scripts from inside the container.

我也是用docker来运行的，最后生成的同样是空文件夹，跟你的问题一样，你现在解决这个问题了嘛？

I have met the same problem , did you solve it

[yipka]

same problem, empty folder

[TitaniumOne]

@WangShuu @a-whitej @tz0403 你好，我也是docker运行生成空文件夹，请问你们是怎么解决的呀？谢谢！！

[Yuuuumie]

@WangShuu @a-whitej @tz0403 你好，我也是docker运行生成空文件夹，请问你们是怎么解决的呀？谢谢！！
你解决了吗？