bob issueshttps://gitlab.idiap.ch/groups/bob/-/issues2018-06-03T12:14:16Zhttps://gitlab.idiap.ch/bob/bob.bio.base/-/issues/27Adding score distribution feature to evaluate.py2018-06-03T12:14:16ZAndré AnjosAdding score distribution feature to evaluate.py*Created by: akomaty*
evaluate.py doesn't have a built-in "plot_score_distribution" feature.
I'll be adding this feature to bob.bio.base asap!*Created by: akomaty*
evaluate.py doesn't have a built-in "plot_score_distribution" feature.
I'll be adding this feature to bob.bio.base asap!https://gitlab.idiap.ch/bob/bob.bio.spear/-/issues/8Mobio database is missing zt-norm2017-10-25T00:08:37ZAndré AnjosMobio database is missing zt-norm*Created by: 183amir*
I tried running some experiments with `mobio-audio-male` with the `--zt-norm` option in `verify_gmm.py` but I realized the database does not support ZT normalization. Is there a reason for that?*Created by: 183amir*
I tried running some experiments with `mobio-audio-male` with the `--zt-norm` option in `verify_gmm.py` but I realized the database does not support ZT normalization. Is there a reason for that?https://gitlab.idiap.ch/bob/bob.bio.spear/-/issues/6unpack requires a string argument of length 4 - broken wave files?2017-10-25T00:08:37ZAndré Anjosunpack requires a string argument of length 4 - broken wave files?*Created by: 1kastner*
I guess the problem is related to a damaged wave files. I downloaded voxforge but because of difficulties I guess some of them are broken. However, I can't see with this error message which part is broken.
marv...*Created by: 1kastner*
I guess the problem is related to a damaged wave files. I downloaded voxforge but because of difficulties I guess some of them are broken. However, I can't see with this error message which part is broken.
marvin@marvin-VirtualBox:~/Downloads/bob-2.0.6$ python ~/Downloads/bob-2.0.6/bin/verify.py -d voxforge -p energy-2gauss -e mfcc-60 -a gmm-voxforge -s ubm_gmm --groups {dev,eval}
/usr/lib/python2.7/dist-packages/scipy/io/wavfile.py:172: WavFileWarning: Chunk (non-data) not understood, skipping it.
WavFileWarning)
bob.bio.base@2015-12-27 22:26:29,894 -- ERROR: During the execution, an exception was raised: unpack requires a string argument of length 4
Traceback (most recent call last):
File "/home/marvin/Downloads/bob-2.0.6/bin/verify.py", line 56, in <module>
sys.exit(bob.bio.base.script.verify.main())
File "/home/marvin/Downloads/bob-2.0.6/eggs/bob.bio.base-2.0.4-py2.7.egg/bob/bio/base/script/verify.py", line 435, in main
verify(args, command_line_parameters)
File "/home/marvin/Downloads/bob-2.0.6/eggs/bob.bio.base-2.0.4-py2.7.egg/bob/bio/base/script/verify.py", line 415, in verify
if not execute(args):
File "/home/marvin/Downloads/bob-2.0.6/eggs/bob.bio.base-2.0.4-py2.7.egg/bob/bio/base/script/verify.py", line 255, in execute
force = args.force)
File "/home/marvin/Downloads/bob-2.0.6/eggs/bob.bio.base-2.0.4-py2.7.egg/bob/bio/base/tools/preprocessor.py", line 59, in preprocess
data = preprocessor.read_original_data(file_name)
File "/home/marvin/Downloads/bob-2.0.6/eggs/bob.bio.spear-2.0.4-py2.7.egg/bob/bio/spear/preprocessor/Base.py", line 21, in read_original_data
rate, audio = scipy.io.wavfile.read(original_file_name)
File "/usr/lib/python2.7/dist-packages/scipy/io/wavfile.py", line 173, in read
_skip_unknown_chunk(fid)
File "/usr/lib/python2.7/dist-packages/scipy/io/wavfile.py", line 90, in _skip_unknown_chunk
size = struct.unpack(fmt, data)[0]
struct.error: unpack requires a string argument of length 4https://gitlab.idiap.ch/bob/bob.bio.base/-/issues/26Confusing defaults documentation on bob.bio.base scripts2018-06-03T12:14:16ZAndré AnjosConfusing defaults documentation on bob.bio.base scripts*Created by: anjos*
Example:
```text
Directories that can be changed according to your requirements:
-T DIR, --temp-directory DIR
The directory for temporary files, default is:
/i...*Created by: anjos*
Example:
```text
Directories that can be changed according to your requirements:
-T DIR, --temp-directory DIR
The directory for temporary files, default is:
/idiap/temp/aanjos/database-name/sub-directory.
(default: None)
-R DIR, --result-directory DIR
The directory for resulting score files, default is:
/idiap/user/aanjos/database-name/sub-directory.
(default: None)
```
What is the right "default"?Manuel Günthersiebenkopf@googlemail.comManuel Günthersiebenkopf@googlemail.comhttps://gitlab.idiap.ch/bob/bob.bio.spear/-/issues/5Minor typo in documentation2017-10-25T00:08:37ZAndré AnjosMinor typo in documentation*Created by: 1kastner*
I hope I am right here when I comment on the documentation.
At http://pythonhosted.org/bob.bio.spear/baselines.html it is suggested to run the following
bin/verify.py -d voxforge -p energy-2gauss -e mfcc_6...*Created by: 1kastner*
I hope I am right here when I comment on the documentation.
At http://pythonhosted.org/bob.bio.spear/baselines.html it is suggested to run the following
bin/verify.py -d voxforge -p energy-2gauss -e mfcc_60 -a gmm-voxforge -s ubm_gmm --groups {dev,eval}
However, the algorithm is registered under mfcc-60.https://gitlab.idiap.ch/bob/bob.bio.spear/-/issues/4Using stereo wave files fails2020-10-08T07:14:21ZAndré AnjosUsing stereo wave files fails*Created by: 1kastner*
First of all great thanks for this existing project. I just tried out some of the features at https://github.com/1kastner/SpeakerRecognition and it worked very nicely. The only thing which bothered me was that I t...*Created by: 1kastner*
First of all great thanks for this existing project. I just tried out some of the features at https://github.com/1kastner/SpeakerRecognition and it worked very nicely. The only thing which bothered me was that I tried to use a stereo file and the error message was quite confusing because it was related to the structure of the numpy array. What would rather follow your conventions: throw an exception with an appropriate error message or just use either left or right and provide the user with a warning? I think making shure the data is valid should be part of the base processing.https://gitlab.idiap.ch/bob/bob.bio.base/-/issues/25It should be possible to list only resources attached to a particular bob.bio...2018-06-03T12:14:16ZAndré AnjosIt should be possible to list only resources attached to a particular bob.bio.XYZ implementation*Created by: anjos*
As of today, all installed resources are listed on the screen.
If a user, for example, is going to run face recognition experiments, but bob.bio.spear is installed on the environment, all resources with for both p...*Created by: anjos*
As of today, all installed resources are listed on the screen.
If a user, for example, is going to run face recognition experiments, but bob.bio.spear is installed on the environment, all resources with for both packages will be listed.
It would be nice to have a command-line option on `resources.py` to filter out stuff that you're not interested on. For example:
```sh
$ ./bin/resources.py -dt database --package=bob.bio.face
```
Would, in this case, list only the resources for the `bob.bio.face` package.https://gitlab.idiap.ch/bob/bob.bio.spear/-/issues/3Cannot reuse existing results from previous experiments2017-10-25T00:08:37ZAndré AnjosCannot reuse existing results from previous experiments*Created by: sun8608*
Sorry to interrupt again! I was trying to reuse results from previous experiments--the preprocessed data:
sudo /usr/local/bin/verify_ivector.py -vv -d voxforge --skip-preprocessing --preprocessed-directory /usr/l...*Created by: sun8608*
Sorry to interrupt again! I was trying to reuse results from previous experiments--the preprocessed data:
sudo /usr/local/bin/verify_ivector.py -vv -d voxforge --skip-preprocessing --preprocessed-directory /usr/local/lib/python2.7/dist-packages/bob.db.voxforge-2.0.2-py2.7.egg/bob/db/voxforge/lists/preprocessed -e mfcc-60 -s ivector_test --groups {dev,eval}
But the "--skip-preprocessing" seems not work and it still requires me to put preprocessor, I follows instructions from the manual, please tell me where I am doing wrong here, thank you! https://gitlab.idiap.ch/bob/bob.bio.spear/-/issues/2Can I execute the bob spear experiment pipelines on files staying in a remote...2017-10-25T00:08:37ZAndré AnjosCan I execute the bob spear experiment pipelines on files staying in a remote server?*Created by: sun8608*
As it states, I apologize if it's not the right place to ask this question. But I think I need to run experiments on files which must stay on the remote server, and meanwhile the problem is bob spear packages canno...*Created by: sun8608*
As it states, I apologize if it's not the right place to ask this question. But I think I need to run experiments on files which must stay on the remote server, and meanwhile the problem is bob spear packages cannot be installed on that server. https://gitlab.idiap.ch/bob/bob.bio.base/-/issues/23Tests failing on the nightlies2018-06-03T12:14:16ZAndré AnjosTests failing on the nightlies*Created by: anjos*
This package is failing on all our Python 3.4 builds with the following error:
```text
======================================================================
ERROR: bob.bio.base.test.test_scripts.test_resources
...*Created by: anjos*
This package is failing on all our Python 3.4 builds with the following error:
```text
======================================================================
ERROR: bob.bio.base.test.test_scripts.test_resources
----------------------------------------------------------------------
Traceback (most recent call last):
File "/opt/local/Library/Frameworks/Python.framework/Versions/3.4/lib/python3.4/site-packages/nose/case.py", line 198, in runTest
self.test(*self.arg)
File "/Users/buildbot/work/gitlab-ci/work/gitlab-ci/f6f95b82/1/biometric/bob.nightlies/layers/2/bob.bio.base/bob/bio/base/test/test_scripts.py", line 373, in test_resources
databases([])
File "/Users/buildbot/work/gitlab-ci/work/gitlab-ci/f6f95b82/1/biometric/bob.nightlies/layers/2/bob.bio.base/bob/bio/base/script/resources.py", line 60, in databases
databases = bob.bio.base.utils.resources.database_directories(replacements=args.database_directories_file)
File "/Users/buildbot/work/gitlab-ci/work/gitlab-ci/f6f95b82/1/biometric/bob.nightlies/layers/2/bob.bio.base/bob/bio/base/utils/resources.py", line 227, in database_directories
for entry_point in sorted(entry_points):
nose.proxy.TypeError: unorderable types: EntryPoint() < EntryPoint()
-------------------- >> begin captured logging << --------------------
```
Ref: https://gitlab.idiap.ch/biometric/bob.nightlies/builds/3106
Not sure how to change this one. Could you please check?https://gitlab.idiap.ch/bob/bob.bio.video/-/issues/3Failing nightly builds2017-10-21T16:05:52ZAndré AnjosFailing nightly builds*Created by: anjos*
Possibly, after the change of `bob.ip.flandmark` to suppress the opencv dependence this package (and also bob.bio.face) require new references for their failing tests. Example:
https://gitlab.idiap.ch/biometric/bo...*Created by: anjos*
Possibly, after the change of `bob.ip.flandmark` to suppress the opencv dependence this package (and also bob.bio.face) require new references for their failing tests. Example:
https://gitlab.idiap.ch/biometric/bob.nightlies/builds/3077
Could you please fix this and push it into the nightlies? Thanks.https://gitlab.idiap.ch/bob/bob.bio.base/-/issues/22Tests are failing on Ubuntu 12.042018-06-03T12:14:17ZAndré AnjosTests are failing on Ubuntu 12.04*Created by: tiagofrepereira2012*
Unit tests are failing on Ubuntu 12.04 (https://gitlab.idiap.ch/biometric/bob.nightlies/builds/2968)
Basically the function `numpy.full` does not exists in numpy 1.6 (the default for the ubuntu 12.04)....*Created by: tiagofrepereira2012*
Unit tests are failing on Ubuntu 12.04 (https://gitlab.idiap.ch/biometric/bob.nightlies/builds/2968)
Basically the function `numpy.full` does not exists in numpy 1.6 (the default for the ubuntu 12.04). The function is available since numpy 1.8.
For example, I would suggest to replace the code
```
numpy.full((1,len(probes)), numpy.nan, numpy.float64)
```
to
```
numpy.ones((1,len(probes)), numpy.float64) * numpy.nan
```
https://gitlab.idiap.ch/bob/bob.bio.base/-/issues/19Add the IDIAP SGE GPU queue2018-06-03T12:14:17ZAndré AnjosAdd the IDIAP SGE GPU queue*Created by: tiagofrepereira2012*
It is necessary to add specific queues for the IDIAP SGE GPU queues.
I already have them in my fork, I will just sync.*Created by: tiagofrepereira2012*
It is necessary to add specific queues for the IDIAP SGE GPU queues.
I already have them in my fork, I will just sync.https://gitlab.idiap.ch/bob/bob.bio.base/-/issues/16Distance.py using an undeclared variable2018-06-03T12:14:17ZAndré AnjosDistance.py using an undeclared variable*Created by: tiagofrepereira2012*
The commit e26e2727abbbf6fa1875215ace2b9e3e2edd85d3 is buggy.
It is using an undeclared variable.
Fix is on the way*Created by: tiagofrepereira2012*
The commit e26e2727abbbf6fa1875215ace2b9e3e2edd85d3 is buggy.
It is using an undeclared variable.
Fix is on the wayhttps://gitlab.idiap.ch/bob/bob.bio.video/-/issues/2AttributeError: 'bob.learn.em.GMMStats' object has no attribute 'copy'2017-10-21T16:05:52ZAndré AnjosAttributeError: 'bob.learn.em.GMMStats' object has no attribute 'copy'*Created by: 183amir*
I am trying to run an experiment with `bob.bio.video` and a `gmm` algorithm and I am seeing this error:
```
<Job: 4 (1281753) [1-32:1] - 'project'> | q_1day : failure (1) -- 'antispoofing.experiment/bin/verify.py...*Created by: 183amir*
I am trying to run an experiment with `bob.bio.video` and a `gmm` algorithm and I am seeing this error:
```
<Job: 4 (1281753) [1-32:1] - 'project'> | q_1day : failure (1) -- 'antispoofing.experiment/bin/verify.py' -vv -d 'replay-licit' -p 'bob.bio.video.preprocessor.Wrapper(bob
.bio.face.preprocessor.TanTriggs(bob.bio.face.preprocessor.FaceCrop(cropped_image_size=(80,64),cropped_positions={"topleft":(0,0),"bottomright":(80,64)})))' -e 'bob.bio.video.extractor.Wrapper("dct-blocks")' -a
'bob.bio.video.algorithm.Wrapper("gmm")' -s 'replay-licit' --groups 'dev' 'eval' --grid 'demanding' -I 'bob.bio.base' 'bob.bio.video' 'bob.bio.face' --sub-task 'project'
Array Job 1 (helvetix10) :
bob.bio.base@2016-05-17 19:58:48,292 -- INFO: - Projection: splitting of index range (0, 10)
bob.bio.base@2016-05-17 19:58:48,293 -- INFO: - Projection: projecting 10 features from directory '/idiap_temp/amohammadi/replay/replay-licit/./extracted' to directory '/idiap_temp/amohammadi/replay/replay-licit
/./projected'
--------------------
bob.bio.base@2016-05-17 19:58:40,133 -- WARNING: Under the desired name 'demanding', there are multiple entry points defined, we return the first one: ['bob.bio.base.config.grid.demanding', 'bob.bio.spear.config
.grid.demanding']
bob.bio.base@2016-05-17 19:58:51,275 -- ERROR: During the execution, an exception was raised: 'bob.learn.em.GMMStats' object has no attribute 'copy'
Traceback (most recent call last):
File "antispoofing.experiment/bin/verify.py", line 29, in <module>
sys.exit(bob.bio.base.script.verify.main())
File "/idiap/bob-stable/py27/local/lib/python2.7/site-packages/bob/bio/base/script/verify.py", line 435, in main
verify(args, command_line_parameters)
File "/idiap/bob-stable/py27/local/lib/python2.7/site-packages/bob/bio/base/script/verify.py", line 382, in verify
if not execute(args):
File "/idiap/bob-stable/py27/local/lib/python2.7/site-packages/bob/bio/base/script/verify.py", line 287, in execute
force = args.force)
File "/idiap/bob-stable/py27/local/lib/python2.7/site-packages/bob/bio/base/tools/algorithm.py", line 116, in project
projected = algorithm.project(feature)
File "antispoofing.experiment/src/bob.bio.video/bob/bio/video/algorithm/Wrapper.py", line 127, in project
features = projected if isinstance(projected, (list,tuple)) else projected.copy()
AttributeError: 'bob.learn.em.GMMStats' object has no attribute 'copy'
```https://gitlab.idiap.ch/bob/bob.bio.spear/-/issues/1Cannot run a trial experiment2017-10-25T00:08:37ZAndré AnjosCannot run a trial experiment*Created by: sun8608*
Hello! I was just trying to run a trial experiment using Voxforge database following the instruction of manuals: sudo /usr/local/bin/verify.py -vv -d voxforge -p energy-2gauss -e mfcc-60 -a gmm-voxforge -s ubm_gmm ...*Created by: sun8608*
Hello! I was just trying to run a trial experiment using Voxforge database following the instruction of manuals: sudo /usr/local/bin/verify.py -vv -d voxforge -p energy-2gauss -e mfcc-60 -a gmm-voxforge -s ubm_gmm --groups {dev,eval}
but I kept getting the error message:
bob.bio.base@2015-08-01 11:03:15,971 -- INFO: - Preprocessing: processing 6561 data files from directory '[YOUR_VOXFORGE_WAV_DIRECTORY]' to directory 'temp/ubm_gmm/preprocessed'
bob.bio.base@2015-08-01 11:03:15,975 -- ERROR: During the execution, an exception was raised: [Errno 2] No such file or directory: '[YOUR_VOXFORGE_WAV_DIRECTORY]/ColinBeckingham-20091029-hfv/wav/a0001.wav'
Traceback (most recent call last):
File "/usr/local/bin/verify.py", line 9, in <module>
load_entry_point('bob.bio.base==2.0.1', 'console_scripts', 'verify.py')()
File "build/bdist.linux-x86_64/egg/bob/bio/base/script/verify.py", line 435, in main
File "build/bdist.linux-x86_64/egg/bob/bio/base/script/verify.py", line 415, in verify
File "build/bdist.linux-x86_64/egg/bob/bio/base/script/verify.py", line 255, in execute
File "build/bdist.linux-x86_64/egg/bob/bio/base/tools/preprocessor.py", line 58, in preprocess
File "build/bdist.linux-x86_64/egg/bob/bio/spear/preprocessor/Base.py", line 21, in read_original_data
File "/usr/lib/python2.7/dist-packages/scipy/io/wavfile.py", line 108, in read
fid = open(file, 'rb')
IOError: [Errno 2] No such file or directory: '[YOUR_VOXFORGE_WAV_DIRECTORY]/ColinBeckingham-20091029-hfv/wav/a0001.wav'
I do not know where I can change the config file, I cannot find a folder with name 'config', there is one 'config' folder in under bob spear installation directory, but changing configuration of this one did not solve my problem; also the directory--build/bdist.linux-x86_64/ is actually empty, I do not know how the messages are retrieved.
https://gitlab.idiap.ch/bob/bob.bio.face/-/issues/6Failing nightly builds2017-10-20T02:30:36ZAndré AnjosFailing nightly builds*Created by: anjos*
Possibly, after the change of `bob.ip.flandmark` to suppress the opencv dependence this package (and also bob.bio.video) require new references for their failing tests. Example:
https://gitlab.idiap.ch/biometric/b...*Created by: anjos*
Possibly, after the change of `bob.ip.flandmark` to suppress the opencv dependence this package (and also bob.bio.video) require new references for their failing tests. Example:
https://gitlab.idiap.ch/biometric/bob.nightlies/builds/3077
Could you please fix this and push it into the nightlies? Thanks.https://gitlab.idiap.ch/bob/bob.bio.base/-/issues/14The script verify.py needs an argument to pass environment variables to the job2018-06-03T12:14:18ZAndré AnjosThe script verify.py needs an argument to pass environment variables to the job*Created by: tiagofrepereira2012*
gridtk jman has this feature.
It is possible to set environment variables to the grid submission using the argument `--environment`.
Would be useful to have the same feature in the `verify.py` script...*Created by: tiagofrepereira2012*
gridtk jman has this feature.
It is possible to set environment variables to the grid submission using the argument `--environment`.
Would be useful to have the same feature in the `verify.py` script.
A patch is on the wayhttps://gitlab.idiap.ch/bob/bob.bio.base/-/issues/13score fusion script lacks documentation and testing2018-06-03T12:14:18ZAndré Anjosscore fusion script lacks documentation and testing*Created by: siebenkopf*
There is a script ``./bin/fusion_llr.py`` that is not documented anywhere, and it seems to be untested. Fix that!*Created by: siebenkopf*
There is a script ``./bin/fusion_llr.py`` that is not documented anywhere, and it seems to be untested. Fix that!https://gitlab.idiap.ch/bob/bob.bio.face/-/issues/3Travis pip install --upgrade numpy may not be the best2017-10-20T02:30:36ZAndré AnjosTravis pip install --upgrade numpy may not be the best*Created by: anjos*
I have found that this package does a `pip install --upgrade numpy` (and other packages). For those standard packages, I think running with `--upgrade` is not a good choice.
We have wheels that are pre-compiled ag...*Created by: anjos*
I have found that this package does a `pip install --upgrade numpy` (and other packages). For those standard packages, I think running with `--upgrade` is not a good choice.
We have wheels that are pre-compiled against the numpy version which is already pre-installed on Travis virtualenvs. If you issue an `--upgrade`, it may create conflicts between the numpy version installed and that with which our SciPy wheel was compiled against.
As a suggestion, I'd completely remove this line and let the requirements to be satisfied (automatically) with the next pip installs which follow this line. If that does not work because of test requirements, I'd setup a "test.txt" requirements file and do a simple `pip install <our-idiap-wheel-repo> -r text.txt` to make sure those are installed properly.
I'm not sure other packages are the same, but we should apply the same strategy everywhere.