bob.bio.face

37 views
Skip to first unread message

nora

unread,
Jan 2, 2018, 1:13:31 PM1/2/18
to bob-devel
Happy new year and Good day everyone:

I was trying to run the baseline experiment using the information from  the bob.bio.face . When I run the command
baselines.py I get the following message:

(bob_py3) peter@peter-Satellite-C850D:~/Downloads$ baselines.py
Could not load the algorithms from the CSU resources. Did you specify bob.bio.csu in your config file?

then I did this:

pip install bob.bio.csu and I had many things but I noticed some information in red ,I listed them below:
==
Failed building wheel for pillow
==
In file included from _imagingtk.c:19:0:
/home/peter/.conda/envs/bob_py3/include/tk.h:96:25: fatal error: X11/Xlib.h: No such file or directory
compilation terminated.
In file included from Tk/tkImaging.c:52:0:
/home/peter/.conda/envs/bob_py3/include/tk.h:96:25: fatal error: X11/Xlib.h: No such file or directory
compilation terminated.
gcc -pthread -shared -B /home/peter/.conda/envs/bob_py3/compiler_compat -L/home/peter/.conda/envs/bob_py3/lib -Wl,-rpath=/home/peter/.conda/envs/bob_py3/lib -Wl,--no-as-needed -Wl,--sysroot=/ build/temp.linux-x86_64-3.6/_imagingtk.o build/temp.linux-x86_64-3.6/Tk/tkImaging.o -L/home/peter/.conda/envs/bob_py3/lib -L/lib64 -L/usr/lib/x86_64-linux-gnu -L/usr/local/lib -L/usr/lib -L/usr/lib/x86_64-linux-gnu -ltcl8.6 -ltk8.6 -o build/lib.linux-x86_64-3.6/PIL/_imagingtk.cpython-36m-x86_64-linux-gnu.so
gcc: error: build/temp.linux-x86_64-3.6/_imagingtk.o: No such file or directory
gcc: error: build/temp.linux-x86_64-3.6/Tk/tkImaging.o: No such file or directory
error: command 'gcc' failed with exit status 1

----------------------------------------
Rolling back uninstall of Pillow
Command "/home/peter/.conda/envs/bob_py3/bin/python -u -c "import setuptools, tokenize;__file__='/tmp/pip-build-q900rhp9/pillow/setup.py';f=getattr(tokenize, 'open', open)(__file__);code=f.read().replace('\r\n', '\n');f.close();exec(compile(code, __file__, 'exec'))" install --record /tmp/pip-w6mmg0s1-record/install-record.txt --single-version-externally-managed --compile" failed with error code 1 in /tmp/pip-build-q900rhp9/pillow/
==

Could you please advise on how to proceed ? Thanks

Manuel Günther

unread,
Jan 2, 2018, 2:07:36 PM1/2/18
to bob-devel
Dear Nora,

the message that you saw is not an error message and, hence, bob.bio.csu is not required to run the baseline algorithms. Just ignore the message.

The bob.bio.csu package is a wrapper package for the CSU Face Recognition Resources.
I believe that we currently have not changed the package to our conda installation and, hence, you might not be able to use it.
If you want to try to install and use the package, you might want to follow the installation instructions here: https://www.idiap.ch/software/bob/docs/bob/bob.bio.csu/master/installation.html (which are, unfortunately, not up-to-date), but there is no guarantee that this will work.
But as mentioned above, you do not need the package at all.

Note to myself and the Idiapers: We should remove this message since it seems to cause trouble. 
It would also be great to port bob.bio.csu to the conda packaging, but I don't know how to deal with the external dependency (CSU Face Recognition Resources), which we are not allowed to redistribute.

Manuel

nora

unread,
Jan 2, 2018, 2:32:06 PM1/2/18
to bob-devel
Thanks a lot Manuel for the information.I ignored this message.after I run baselines.py I can see the folder baselines in my temp and result file,after that I tried the command

baselines.py --all -vv --evaluate ROC DET CMC HTER

I did not get any figure and here is what I got

=====
(bob_py3) peter@peter-Satellite-C850D:~/Downloads$ baselines.py --all -vv --evaluate ROC DET CMC HTER

Could not load the algorithms from the CSU resources. Did you specify bob.bio.csu in your config file?
bob.bio.face@2018-01-02 12:20:44,294 -- WARNING: Skipping algorithm 'lda' since the results cannot be found.
bob.bio.face@2018-01-02 12:20:44,295 -- WARNING: Skipping algorithm 'gabor-graph' since the results cannot be found.
bob.bio.face@2018-01-02 12:20:44,296 -- WARNING: Skipping algorithm 'lgbphs' since the results cannot be found.
bob.bio.face@2018-01-02 12:20:44,296 -- WARNING: Skipping algorithm 'plda' since the results cannot be found.
bob.bio.face@2018-01-02 12:20:44,296 -- WARNING: Skipping algorithm 'bic' since the results cannot be found.
bob.bio.face@2018-01-02 12:20:44,297 -- WARNING: Skipping algorithm 'gmm' since the results cannot be found.
bob.bio.face@2018-01-02 12:20:44,297 -- WARNING: Skipping algorithm 'isv' since the results cannot be found.
bob.bio.face@2018-01-02 12:20:44,297 -- WARNING: Skipping algorithm 'ivector' since the results cannot be found.
bob.bio.face@2018-01-02 12:20:44,323 -- INFO: Executing command (nonorm):
/home/peter/.conda/envs/bob_py3/bin/evaluate.py --directory 'results/baselines' --legends 'eigenface' --criterion 'HTER' --roc 'ROC.pdf' --det 'DET.pdf' --cmc 'CMC.pdf' -vv --dev-files 'eigenface/Default/nonorm/scores-dev'
Traceback (most recent call last):
  File "/home/peter/.conda/envs/bob_py3/bin/evaluate.py", line 6, in <module>
    sys.exit(bob.bio.base.script.evaluate.main())
  File "/home/peter/.conda/envs/bob_py3/lib/python3.6/site-packages/bob/bio/base/script/evaluate.py", line 277, in main
    args = command_line_arguments(command_line_parameters)
  File "/home/peter/.conda/envs/bob_py3/lib/python3.6/site-packages/bob/bio/base/script/evaluate.py", line 79, in command_line_arguments
    raise ValueError("The provided score file '%s' does not exist", f)
ValueError: ("The provided score file '%s' does not exist", 'eigenface/Default/nonorm/scores-dev')
==
do i need to add the path to scores-dev ?  how about the other parameters,do I have to specify them ?   Thanks a lot

Manuel Günther

unread,
Jan 2, 2018, 8:34:36 PM1/2/18
to bob-devel
Nora,

I believe this is due to an already fixed bug, via https://gitlab.idiap.ch/bob/bob.bio.base/merge_requests/108. Apparently, we did not publish a new version of bob.bio.base yet, or you do not have the latest version of bob.bio.base.

@Idiapers: I have pushed a new MR for the wrong error message in the ValueError you can see above. Could you please publish a new version of bob.bio.base in conda, after merging that?

Thanks
Manuel

Tiago Freitas Pereira

unread,
Jan 3, 2018, 2:45:33 AM1/3/18
to bob-...@googlegroups.com
Hi Manuel,

I've tried to keep this package under our nightly builds without success as you can see here https://gitlab.idiap.ch/bob/bob.bio.csu/issues/1.
In fact the CSU package needs a maintenance, since the last release was in 2012 (http://www.cs.colostate.edu/facerec/algorithms/baselines2011.php).
@Manuel and @Idiapers, is it reasonable to archive bob.bio.csu package?

About the release of bob.bio.base, we need to release a new version of the whole Bob since the last one was 3 months ago. What do you think @Idiapers?

Cheers




--
-- You received this message because you are subscribed to the Google Groups bob-devel group. To post to this group, send email to bob-...@googlegroups.com. To unsubscribe from this group, send email to bob-devel+unsubscribe@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/bob-devel or directly the project website at http://idiap.github.com/bob/
---
You received this message because you are subscribed to the Google Groups "bob-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email to bob-devel+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.



--
Tiago

Amir Mohammadi

unread,
Jan 3, 2018, 8:12:03 AM1/3/18
to bob-...@googlegroups.com
Hi,

The new release is being blocked by https://gitlab.idiap.ch/bob/bob.admin/issues/35 which I just updated.
We need some help with that issue to move forward. It is an ongoing effort.

contain information on how to use bob packages from source.

Best,
Amir

On Wed, Jan 3, 2018 at 8:45 AM Tiago Freitas Pereira <tiagofr...@gmail.com> wrote:
Hi Manuel,

I've tried to keep this package under our nightly builds without success as you can see here https://gitlab.idiap.ch/bob/bob.bio.csu/issues/1.
In fact the CSU package needs a maintenance, since the last release was in 2012 (http://www.cs.colostate.edu/facerec/algorithms/baselines2011.php).
@Manuel and @Idiapers, is it reasonable to archive bob.bio.csu package?

About the release of bob.bio.base, we need to release a new version of the whole Bob since the last one was 3 months ago. What do you think @Idiapers?

Cheers



On Wed, Jan 3, 2018 at 2:34 AM, 'Manuel Günther' via bob-devel <bob-...@googlegroups.com> wrote:
Nora,

I believe this is due to an already fixed bug, via https://gitlab.idiap.ch/bob/bob.bio.base/merge_requests/108. Apparently, we did not publish a new version of bob.bio.base yet, or you do not have the latest version of bob.bio.base.

@Idiapers: I have pushed a new MR for the wrong error message in the ValueError you can see above. Could you please publish a new version of bob.bio.base in conda, after merging that?

Thanks
Manuel

--
-- You received this message because you are subscribed to the Google Groups bob-devel group. To post to this group, send email to bob-...@googlegroups.com. To unsubscribe from this group, send email to bob-devel+...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/bob-devel or directly the project website at http://idiap.github.com/bob/

---
You received this message because you are subscribed to the Google Groups "bob-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email to bob-devel+...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.



--
Tiago

--
-- You received this message because you are subscribed to the Google Groups bob-devel group. To post to this group, send email to bob-...@googlegroups.com. To unsubscribe from this group, send email to bob-devel+...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/bob-devel or directly the project website at http://idiap.github.com/bob/

---
You received this message because you are subscribed to the Google Groups "bob-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email to bob-devel+...@googlegroups.com.

nora

unread,
Jan 3, 2018, 11:47:25 AM1/3/18
to bob-devel
Hi:

I tried to install bob.bio.base again
==
(bob_py3) peter@peter-Satellite-C850D:~/Downloads$ conda install bob.bio.base
Fetching package metadata .............
Solving package specifications: .

# All requested packages already installed.
# packages in environment at /home/peter/.conda/envs/bob_py3:
#
bob.bio.base              3.2.1                      py_0    https://www.idiap.ch/software/bob/conda
==

and tried again the command below

 baselines.py --all -vv --evaluate ROC DET CMC HTER

I am still getting the same error message,I will appreciate any help on how to proceed.

Amir Mohammadi

unread,
Jan 3, 2018, 12:06:17 PM1/3/18
to bob-...@googlegroups.com
Nora,

Your error says the score files are not found!
Have you looked for them? Are they created? where are they?
Try to use evaluate.py for an alternative please.

Thanks,
Amir

--

nora

unread,
Jan 3, 2018, 1:00:03 PM1/3/18
to bob-devel
Hi Amir:

yes i do have the dev-scores,I tried to add the path of the dev-scores to the baselines.py command and it is still giving me the same messages as below:

==
(bob_py3) peter@peter-Satellite-C850D:~/Downloads$ baselines.py --all -vv -b /home/peter/Downloads/results/baselines --dry-run -e EER

Could not load the algorithms from the CSU resources. Did you specify bob.bio.csu in your config file?
bob.bio.face@2018-01-03 10:27:58,897 -- WARNING: Skipping algorithm 'lda' since the results cannot be found.
bob.bio.face@2018-01-03 10:27:58,899 -- WARNING: Skipping algorithm 'gabor-graph' since the results cannot be found.
bob.bio.face@2018-01-03 10:27:58,899 -- WARNING: Skipping algorithm 'lgbphs' since the results cannot be found.
bob.bio.face@2018-01-03 10:27:58,900 -- WARNING: Skipping algorithm 'plda' since the results cannot be found.
bob.bio.face@2018-01-03 10:27:58,900 -- WARNING: Skipping algorithm 'bic' since the results cannot be found.
bob.bio.face@2018-01-03 10:27:58,900 -- WARNING: Skipping algorithm 'gmm' since the results cannot be found.
bob.bio.face@2018-01-03 10:27:58,901 -- WARNING: Skipping algorithm 'isv' since the results cannot be found.
bob.bio.face@2018-01-03 10:27:58,901 -- WARNING: Skipping algorithm 'ivector' since the results cannot be found.
bob.bio.face@2018-01-03 10:27:58,927 -- INFO: Executing command (nonorm):
/home/peter/.conda/envs/bob_py3/bin/evaluate.py --directory /home/peter/Downloads/results/baselines --legends 'eigenface' --criterion 'EER' -vv --dev-files 'eigenface/Default/nonorm/scores-dev'
==
I also tried
==
(bob_py3) peter@peter-Satellite-C850D:~/Downloads$ baselines.py --all -vv -b /home/peter/Downloads/results/baselines/eigenface/Default/nonorm/scores-dev --dry-run -e 'ROC'
Could not load the algorithms from the CSU resources. Did you specify bob.bio.csu in your config file?
bob.bio.face@2018-01-03 10:38:10,305 -- WARNING: Skipping algorithm 'eigenface' since the results cannot be found.
bob.bio.face@2018-01-03 10:38:10,306 -- WARNING: Skipping algorithm 'lda' since the results cannot be found.
bob.bio.face@2018-01-03 10:38:10,307 -- WARNING: Skipping algorithm 'gabor-graph' since the results cannot be found.
bob.bio.face@2018-01-03 10:38:10,307 -- WARNING: Skipping algorithm 'lgbphs' since the results cannot be found.
bob.bio.face@2018-01-03 10:38:10,307 -- WARNING: Skipping algorithm 'plda' since the results cannot be found.
bob.bio.face@2018-01-03 10:38:10,308 -- WARNING: Skipping algorithm 'bic' since the results cannot be found.
bob.bio.face@2018-01-03 10:38:10,308 -- WARNING: Skipping algorithm 'gmm' since the results cannot be found.
bob.bio.face@2018-01-03 10:38:10,308 -- WARNING: Skipping algorithm 'isv' since the results cannot be found.
bob.bio.face@2018-01-03 10:38:10,309 -- WARNING: Skipping algorithm 'ivector' since the results cannot be found.
bob.bio.face@2018-01-03 10:38:10,334 -- INFO: Executing command (nonorm):
/home/peter/.conda/envs/bob_py3/bin/evaluate.py --directory /home/peter/Downloads/results/baselines/eigenface/Default/nonorm/scores-dev --legends --roc 'ROC.pdf' -vv --dev-files
===
then i used evaluate.py as you suggested and it is working

==
(bob_py3) peter@peter-Satellite-C850D:~/Downloads$ evaluate.py -vv --dev-files results/baselines/eigenface/Default/nonorm/scores-dev --legends eigenface --criterion EER --cmc cmcat.pdf --roc rocat.pdf --det detat.pdf -rr
bob.bio.base@2018-01-03 10:42:51,863 -- INFO: Loading 1 score files of the development set
bob.bio.base@2018-01-03 10:42:51,909 -- INFO: Computing EER on the development set
The EER of the development set of 'eigenface' is 9.000%
bob.bio.base@2018-01-03 10:42:51,911 -- INFO: Computing CAR curves on the development set
bob.bio.base@2018-01-03 10:42:51,912 -- INFO: Plotting ROC curves to file 'rocat.pdf'
bob.bio.base@2018-01-03 10:42:55,424 -- INFO: Computing DET curves on the development set
bob.bio.base@2018-01-03 10:42:55,432 -- INFO: Plotting DET curves to file 'detat.pdf'
bob.bio.base@2018-01-03 10:42:56,556 -- INFO: Loading CMC data on the development set
bob.bio.base@2018-01-03 10:42:56,578 -- INFO: Plotting CMC curves to file 'cmcat.pdf'
bob.bio.base@2018-01-03 10:42:57,847 -- INFO: Computing recognition rate on the development set
The Recognition Rate of the development set of 'eigenface' is 79.000%
==

so will I just use evaluate.py in the future ?  Thanks a lot for your help,appreciate it..

Amir Mohammadi

unread,
Jan 3, 2018, 1:06:02 PM1/3/18
to bob-...@googlegroups.com
I would use evaluate.py yes.

Manuel Günther

unread,
Jan 3, 2018, 1:06:42 PM1/3/18
to bob-devel
Amir, this is a known and already solved bug in bob.bio.base via https://gitlab.idiap.ch/bob/bob.bio.base/merge_requests/108 It is definitely not Nora's fault.

Nora, while we are unable to publish a new version of bob.bio.base at this instance, you can use the docs of bob extension (which Amir pointed to): https://www.idiap.ch/software/bob/docs/bob/bob.extension/master/ to see how to use the current master branch version of bob.bio.base. Let us know if you have any issue with the documentation. You can also use "evaluate.py" by hand, as you have done it, by avoiding to use the "--directory" option (which is buggy in the stable version, and has been fixed in the master branch).

Best regards
Manuel

Amir Mohammadi

unread,
Jan 3, 2018, 1:10:08 PM1/3/18
to bob-...@googlegroups.com

I don't recommend to use the master version of bob.bio.base since you would have to use all other bob packages from source too in theory.


--

nora

unread,
Jan 3, 2018, 1:26:02 PM1/3/18
to bob-devel
Thanks a lot everyone,I looked at the docs of bob extension ,I will keep using evaluate.py for now since it is easier for me :-) . Thanks again
Reply all
Reply to author
Forward
0 new messages