Error parsing text-format caffe.NetParameter: Message type "caffe.LayerParameter" has no field named "bn_param".

[buaa-luzhi]

Hi, @yjxiong
python tools/eval_net.py ucf101 1 rgb FRAME_PATH
models/ucf101/tsn_bn_inception_rgb_deploy.prototxt models/ucf101_split_1_tsn_rgb_reference_bn_inception.caffemodel
--num_worker 4 --save_scores SCORE_FILE

W0901 14:59:26.851496 12575 _caffe.cpp:128] Net('models/hmdb51/tsn_bn_inception_rgb_deploy.prototxt', 1, weights='models/hmdb51_split_1_tsn_rgb_reference_bn_inception.caffemodel')
[libprotobuf ERROR google/protobuf/text_format.cc:245] Error parsing text-format caffe.NetParameter: 19:12: Message type "caffe.LayerParameter" has no field named "bn_param".
F0901 14:59:26.854511 12575 upgrade_proto.cpp:88] Check failed: ReadProtoFromTextFile(param_file, param) Failed to parse NetParameter file: models/hmdb51/tsn_bn_inception_rgb_deploy.prototxt
*** Check failure stack trace: ***
Aborted (core dumped)

caffe-action has been installed successfully.
Can you give me some advice about how to solve it!

Thank you so much!

[yjxiong]

Have you strictly followed the instructions in README? You are running on another version of Caffe installed.

[buaa-luzhi]

yes, according to the ReadMe every step, and download caffe from "https://github.com/yjxiong/caffe" and installed successfully.

[yjxiong]

I don't know which README you are reading. But people usually go step by step starting from

https://github.com/yjxiong/temporal-segment-networks#get-the-code

and follow on.

[buaa-luzhi]

but, i have only one gpu, this is different from you

[buaa-luzhi]

The ReadMe is in the temporal-segment-networks zip file, i download it from "https://github.com/yjxiong"

[yjxiong]

Then I guess it is the same as the one in the front page of this page. Please read it carefully and follow the step by step guide.

[buaa-luzhi]

Thank you!

[yjxiong]

I have said in the first reply you are running another version of Caffe. If you have followed the instructions on how to install, this should not have happened.

[buaa-luzhi]

I'll reinstall caffe following the instructions. thank you!

[buaa-luzhi]

@yjxiong
Sorry to disturb you again!
First，git clone --recursive https://github.com/yjxiong/temporal-segment-networks
Second compile caffe, mkdir build && cd build cmake .. -DUSE_MPI=ON make && make install
Third， dataset has been download
Fourth, python tools/eval_net.py ucf101 1 rgb FRAME_PATH
models/ucf101/tsn_bn_inception_rgb_deploy.prototxt models/ucf101_split_1_tsn_rgb_reference_bn_inception.caffemodel
--num_worker 4 --save_scores SCORE_FILE

Still making mistakes.

I hope you can help me again, thank you!

[yjxiong]

Did you use the build_all.sh script? I am not able to help you if you keep not following the instructions.

[buaa-luzhi]

Because, i have been installed Opencv-3-2-0-rc.
so, if i used build_all.sh, i am worry about conflict with Opencv2..4.13.

[yjxiong]

It simply won't.

[buaa-luzhi]

Do you have any advice, thank you!

[buaa-luzhi]

Ok, thank you, i will try.

[buaa-luzhi]

@yjxiong
Hello,
Thanks for your reply, I am so sorry disturb you again!
First, run the "build_all.sh" script, everything is ok,
However, when execute python tools/evel_net.py ... , this error occurs again.
At the moment, I feel helpless

[yjxiong]

I have said you are running another Caffe installed on your machine. Another, less possible, reason could be that you have libboost version lower than 1.55, which will lead to the Python interface failing to compile.

[buaa-luzhi]

@yjxiong
Yes, you are right, i have solved this problem!
Thank you for your reply!

[yjxiong]

Closing this. Please feel free to reopen it if you meet any further problem.