Problems arise when running ./run.sh
Closed this issue · 31 comments
Hello
There are some problems when I run ./run.sh. And I can't find the way to solve it.
Traceback (most recent call last):
File "compute_detector.py", line 47, in
from Utils.sift_tools import recomputeOrientation
File "/home/hoo104/LIFT/python-code/Utils/sift_tools.py", line 48, in
libSIFT = cdll.LoadLibrary('../c-code/libSIFT.so')
File "/home/hoo104/anaconda3/lib/python3.5/ctypes/init.py", line 425, in LoadLibrary
return self._dlltype(name)
File "/home/hoo104/anaconda3/lib/python3.5/ctypes/init.py", line 347, in init
self._handle = _dlopen(self._name, mode)
OSError: ../c-code/libSIFT.so: undefined symbol: _ZN2cv3hal9fastAtan2EPKfS2_Pfib
Parameters
#--------------------------------------------------------------------------------------
Be careful as we do not use sophisticated parsing
The parser will read until the semicolon
types are defined as below
ss: multiple strings separated with commas
s: string
b: boolean
f: float
d: int
#--------------------------------------------------------------------------------------
#--------------------------------------------------------------------------------------
Dataset parameters
ss: dataset.trainSetList = ECCV/piccadilly/; # All the first images of oxford dataset is used for training
s: dataset.dataType = ECCV; # the dataset type
#--------------------------------------------------------------------------------------
Model parameters
s: model.modelType = Combined; # the network type
b: model.bNormalizeInput = 1; # Normalize input to have zero mean and 1 std
f: model.fScaleList = np.array([1.0]); # list of scale spaces (small means larger scale)
GHH related
f: model.max_strength = -1; # GHH softmax strength (-1 for hard)
Keypoints
s: model.sDetector = tilde; # network architecture for kp
d: model.nFilterSize = 25; # Let's keep the number odd
d: model.nPatchSizeKp = 48; # patch size for kp including moving
# regions we use something smaller
s: model.sKpNonlinearity = None; # use nonlinearity at end
f: model.fScaleList = np.array([1.0]); # list of scale spaces (small means larger scale)
f: model.bias_rnd = 0.0; # random noise added to bias
f: model.epsilon = 1e-8; # epsilon for tilde
Orientation
s: model.sOrientation = cvpr16; # network architecture for orientation
Descriptor
s: model.sDescriptor = descriptor_imported; # network architecture for desc
d: model.nDescInputSize = 64; # Input size to be fed to the descriptor
s: model.descriptor_export_folder = /cvlabdata1/home/trulls-data/kpdesc/torch/export/;
L-> skipped
s: model.descriptor_model = new-CNN3-picc-iter-56k.h5; # network configuration
s: model.descriptor_input = desc-input; # Descriptor input
#--------------------------------------------------------------------------------------
Parameters for patch extraction
automatically determined
f: patch.fMaxScale = np.max(self.model.fScaleList); # asserts make sure this is stored properly
f: patch.fRatioScale = (self.model.nPatchSizeKp / 2.0) / 2.0; # to not resize when scale is 2
d: patch.nPatchSize = np.round(self.model.nDescInputSize * self.patch.fRatioScale / 6.0); # the large patch size for data. the desc will look at ratio scale of 6.0
#--------------------------------------------------------------------------------------
Validation and test time parameters
d: validation.batch_size = 100; # batch size of the implementation
d: validation.nScaleInterval = 4;
L-> skipped
d: validation.nNMSInterval = 2; # number of intervals we look for
L-> skipped
# NMS (3 would mean it goes up one
# octave in case of
# nScaleInterval=2)
#---------------------------------------------------------------------------------
Test Data Module
Traceback (most recent call last):
File "compute_orientation.py", line 107, in
test_data_in = data_module.data_obj(param, image_file_name, kp_file_name)
File "/home/hoo104/LIFT/python-code/Utils/dataset_tools/test.py", line 64, in init
self.load_data(param, image_file_name, kp_file_name)
File "/home/hoo104/LIFT/python-code/Utils/dataset_tools/test.py", line 82, in load_data
pathconf, param, image_file_name, kp_file_name)
File "/home/hoo104/LIFT/python-code/Utils/dataset_tools/test.py", line 129, in load_data_for_set
kp = np.asarray(loadKpListFromTxt(kp_file_name))
File "/home/hoo104/LIFT/python-code/Utils/kp_tools.py", line 196, in loadKpListFromTxt
kp_file = open(kp_file_name, 'rb')
FileNotFoundError: [Errno 2] No such file or directory: '/home/hoo104/LIFT/results/img1_kp.txt'
Parameters
#--------------------------------------------------------------------------------------
Be careful as we do not use sophisticated parsing
The parser will read until the semicolon
types are defined as below
ss: multiple strings separated with commas
s: string
b: boolean
f: float
d: int
#--------------------------------------------------------------------------------------
#--------------------------------------------------------------------------------------
Dataset parameters
ss: dataset.trainSetList = ECCV/piccadilly/; # All the first images of oxford dataset is used for training
s: dataset.dataType = ECCV; # the dataset type
#--------------------------------------------------------------------------------------
Model parameters
s: model.modelType = Combined; # the network type
b: model.bNormalizeInput = 1; # Normalize input to have zero mean and 1 std
f: model.fScaleList = np.array([1.0]); # list of scale spaces (small means larger scale)
GHH related
f: model.max_strength = -1; # GHH softmax strength (-1 for hard)
Keypoints
s: model.sDetector = tilde; # network architecture for kp
d: model.nFilterSize = 25; # Let's keep the number odd
d: model.nPatchSizeKp = 48; # patch size for kp including moving
# regions we use something smaller
s: model.sKpNonlinearity = None; # use nonlinearity at end
f: model.fScaleList = np.array([1.0]); # list of scale spaces (small means larger scale)
f: model.bias_rnd = 0.0; # random noise added to bias
f: model.epsilon = 1e-8; # epsilon for tilde
Orientation
s: model.sOrientation = cvpr16; # network architecture for orientation
Descriptor
s: model.sDescriptor = descriptor_imported; # network architecture for desc
d: model.nDescInputSize = 64; # Input size to be fed to the descriptor
s: model.descriptor_export_folder = /cvlabdata1/home/trulls-data/kpdesc/torch/export/;
L-> skipped
s: model.descriptor_model = new-CNN3-picc-iter-56k.h5; # network configuration
s: model.descriptor_input = desc-input; # Descriptor input
#--------------------------------------------------------------------------------------
Parameters for patch extraction
automatically determined
f: patch.fMaxScale = np.max(self.model.fScaleList); # asserts make sure this is stored properly
f: patch.fRatioScale = (self.model.nPatchSizeKp / 2.0) / 2.0; # to not resize when scale is 2
d: patch.nPatchSize = np.round(self.model.nDescInputSize * self.patch.fRatioScale / 6.0); # the large patch size for data. the desc will look at ratio scale of 6.0
#--------------------------------------------------------------------------------------
Validation and test time parameters
d: validation.batch_size = 100; # batch size of the implementation
d: validation.nScaleInterval = 4;
L-> skipped
d: validation.nNMSInterval = 2; # number of intervals we look for
L-> skipped
# NMS (3 would mean it goes up one
# octave in case of
# nScaleInterval=2)
#---------------------------------------------------------------------------------
Test Data Module
Traceback (most recent call last):
File "compute_descriptor.py", line 111, in
test_data_in = data_module.data_obj(param, image_file_name, kp_file_name)
File "/home/hoo104/LIFT/python-code/Utils/dataset_tools/test.py", line 64, in init
self.load_data(param, image_file_name, kp_file_name)
File "/home/hoo104/LIFT/python-code/Utils/dataset_tools/test.py", line 82, in load_data
pathconf, param, image_file_name, kp_file_name)
File "/home/hoo104/LIFT/python-code/Utils/dataset_tools/test.py", line 135, in load_data_for_set
angle = np.pi / 180.0 * kp[:, IDX_ANGLE] # store angle in radians
IndexError: too many indices for array
@kmyid
Thank you very much.
When I download the new file and compile it again, the problem has disappeared. But there are some others questions,like
1.TypeError: list indices must be integers or slices, not numpy.float64
2.FileNotFoundError: [Errno 2] No such file or directory: '/home/hoo104/LIFT/results/img1_kp.txt'
3.FileNotFoundError: [Errno 2] No such file or directory: '/home/hoo104/LIFT/results/img1_ori.txt'
The first problem seems the matter of python3.x, but I am not sure.
The second and third problem indicate that I need to create two files named img1_kp.txt and img1_ori.txt?
I would be appreciated if you could give me some hints to continue.
hoo104@amax-1080:~/LIFT$ ./run.sh
Parameters
#--------------------------------------------------------------------------------------
Be careful as we do not use sophisticated parsing
The parser will read until the semicolon
types are defined as below
ss: multiple strings separated with commas
s: string
b: boolean
f: float
d: int
#--------------------------------------------------------------------------------------
#--------------------------------------------------------------------------------------
Dataset parameters
ss: dataset.trainSetList = ECCV/piccadilly/; # All the first images of oxford dataset is used for training
s: dataset.dataType = ECCV; # the dataset type
#--------------------------------------------------------------------------------------
Model parameters
s: model.modelType = Combined; # the network type
b: model.bNormalizeInput = 1; # Normalize input to have zero mean and 1 std
f: model.fScaleList = np.array([1.0]); # list of scale spaces (small means larger scale)
GHH related
f: model.max_strength = -1; # GHH softmax strength (-1 for hard)
Keypoints
s: model.sDetector = tilde; # network architecture for kp
d: model.nFilterSize = 25; # Let's keep the number odd
d: model.nPatchSizeKp = 48; # patch size for kp including moving
# regions we use something smaller
s: model.sKpNonlinearity = None; # use nonlinearity at end
f: model.fScaleList = np.array([1.0]); # list of scale spaces (small means larger scale)
f: model.bias_rnd = 0.0; # random noise added to bias
f: model.epsilon = 1e-8; # epsilon for tilde
Orientation
s: model.sOrientation = cvpr16; # network architecture for orientation
Descriptor
s: model.sDescriptor = descriptor_imported; # network architecture for desc
d: model.nDescInputSize = 64; # Input size to be fed to the descriptor
s: model.descriptor_export_folder = /cvlabdata1/home/trulls-data/kpdesc/torch/export/;
L-> skipped
s: model.descriptor_model = new-CNN3-picc-iter-56k.h5; # network configuration
s: model.descriptor_input = desc-input; # Descriptor input
#--------------------------------------------------------------------------------------
Parameters for patch extraction
automatically determined
f: patch.fMaxScale = np.max(self.model.fScaleList); # asserts make sure this is stored properly
f: patch.fRatioScale = (self.model.nPatchSizeKp / 2.0) / 2.0; # to not resize when scale is 2
d: patch.nPatchSize = np.round(self.model.nDescInputSize * self.patch.fRatioScale / 6.0); # the large patch size for data. the desc will look at ratio scale of 6.0
#--------------------------------------------------------------------------------------
Validation and test time parameters
d: validation.batch_size = 100; # batch size of the implementation
d: validation.nScaleInterval = 4;
L-> skipped
d: validation.nNMSInterval = 2; # number of intervals we look for
L-> skipped
# NMS (3 would mean it goes up one
# octave in case of
# nScaleInterval=2)
#---------------------------------------------------------------------------------
Time taken to read and prepare the image is 54.82700000000001 ms
INFO: Testing double scale
resize to test is [ 1.95833333 1.70482819 1.48413914 1.29201816 1.12476714 0.97916667
0.85241409 0.74206957 0.64600908 0.56238357 0.48958333 0.42620705
0.37103478 0.32300454 0.28119178 0.24479167 0.21310352 0.18551739
0.16150227 0.14059589 0.12239583]
scales to test is [ 1. 1.14869835 1.31950791 1.51571657 1.74110113 2.
2.29739671 2.63901582 3.03143313 3.48220225 4. 4.59479342
5.27803164 6.06286627 6.96440451 8. 9.18958684
10.55606329 12.12573253 13.92880901 16. ]
Time taken to resize image is 16.235ms
WARNING (theano.sandbox.cuda): The cuda backend is deprecated and will be removed in the next release (v0.10). Please switch to the gpuarray backend. You can get more information about how to switch at this URL:
https://github.com/Theano/Theano/wiki/Converting-to-the-new-gpu-back-end%28gpuarray%29
/home/hoo104/anaconda3/lib/python3.5/site-packages/theano/sandbox/cuda/init.py:556: UserWarning: Theano flag device=gpu* (old gpu back-end) only support floatX=float32. You have floatX=float64. Use the new gpu back-end with device=cuda* for that value of floatX.
warnings.warn(msg)
Using gpu device 0: GeForce GTX 1080 (CNMeM is disabled, cuDNN 5105)
Traceback (most recent call last):
File "compute_detector.py", line 215, in
image, verbose=False)
File "/home/hoo104/LIFT/python-code/Utils/solvers.py", line 148, in TestImage
myNet = CreateNetwork4Image(pathconf, param, image, verbose=verbose)
File "/home/hoo104/LIFT/python-code/Utils/solvers.py", line 130, in CreateNetwork4Image
verbose=verbose)
File "/home/hoo104/LIFT/python-code/Utils/networks/eccv_combined.py", line 78, in init
verbose=verbose)
File "/home/hoo104/LIFT/python-code/Utils/networks/eccv_base.py", line 121, in init
self.buildLayers(verbose=verbose, **kwargs)
File "/home/hoo104/LIFT/python-code/Utils/networks/eccv_combined.py", line 105, in buildLayers
not bTestWholeImage), verbose=verbose)
File "/home/hoo104/LIFT/python-code/Utils/networks/eccv_combined.py", line 158, in buildLayersKp
detector_module.build(self, idxSiam, verbose)
File "/home/hoo104/LIFT/python-code/Utils/networks/detector/tilde.py", line 212, in build
eps=myNet.config.epsilon)
File "/home/hoo104/LIFT/python-code/Utils/lasagne_tools.py", line 97, in createXYZMapLayer
scale_space_min = fScaleList[new_min_idx]
TypeError: list indices must be integers or slices, not numpy.float64
Parameters
#--------------------------------------------------------------------------------------
Be careful as we do not use sophisticated parsing
The parser will read until the semicolon
types are defined as below
ss: multiple strings separated with commas
s: string
b: boolean
f: float
d: int
#--------------------------------------------------------------------------------------
#--------------------------------------------------------------------------------------
Dataset parameters
ss: dataset.trainSetList = ECCV/piccadilly/; # All the first images of oxford dataset is used for training
s: dataset.dataType = ECCV; # the dataset type
#--------------------------------------------------------------------------------------
Model parameters
s: model.modelType = Combined; # the network type
b: model.bNormalizeInput = 1; # Normalize input to have zero mean and 1 std
f: model.fScaleList = np.array([1.0]); # list of scale spaces (small means larger scale)
GHH related
f: model.max_strength = -1; # GHH softmax strength (-1 for hard)
Keypoints
s: model.sDetector = tilde; # network architecture for kp
d: model.nFilterSize = 25; # Let's keep the number odd
d: model.nPatchSizeKp = 48; # patch size for kp including moving
# regions we use something smaller
s: model.sKpNonlinearity = None; # use nonlinearity at end
f: model.fScaleList = np.array([1.0]); # list of scale spaces (small means larger scale)
f: model.bias_rnd = 0.0; # random noise added to bias
f: model.epsilon = 1e-8; # epsilon for tilde
Orientation
s: model.sOrientation = cvpr16; # network architecture for orientation
Descriptor
s: model.sDescriptor = descriptor_imported; # network architecture for desc
d: model.nDescInputSize = 64; # Input size to be fed to the descriptor
s: model.descriptor_export_folder = /cvlabdata1/home/trulls-data/kpdesc/torch/export/;
L-> skipped
s: model.descriptor_model = new-CNN3-picc-iter-56k.h5; # network configuration
s: model.descriptor_input = desc-input; # Descriptor input
#--------------------------------------------------------------------------------------
Parameters for patch extraction
automatically determined
f: patch.fMaxScale = np.max(self.model.fScaleList); # asserts make sure this is stored properly
f: patch.fRatioScale = (self.model.nPatchSizeKp / 2.0) / 2.0; # to not resize when scale is 2
d: patch.nPatchSize = np.round(self.model.nDescInputSize * self.patch.fRatioScale / 6.0); # the large patch size for data. the desc will look at ratio scale of 6.0
#--------------------------------------------------------------------------------------
Validation and test time parameters
d: validation.batch_size = 100; # batch size of the implementation
d: validation.nScaleInterval = 4;
L-> skipped
d: validation.nNMSInterval = 2; # number of intervals we look for
L-> skipped
# NMS (3 would mean it goes up one
# octave in case of
# nScaleInterval=2)
#---------------------------------------------------------------------------------
Test Data Module
Traceback (most recent call last):
File "compute_orientation.py", line 107, in
test_data_in = data_module.data_obj(param, image_file_name, kp_file_name)
File "/home/hoo104/LIFT/python-code/Utils/dataset_tools/test.py", line 64, in init
self.load_data(param, image_file_name, kp_file_name)
File "/home/hoo104/LIFT/python-code/Utils/dataset_tools/test.py", line 82, in load_data
pathconf, param, image_file_name, kp_file_name)
File "/home/hoo104/LIFT/python-code/Utils/dataset_tools/test.py", line 129, in load_data_for_set
kp = np.asarray(loadKpListFromTxt(kp_file_name))
File "/home/hoo104/LIFT/python-code/Utils/kp_tools.py", line 196, in loadKpListFromTxt
kp_file = open(kp_file_name, 'rb')
FileNotFoundError: [Errno 2] No such file or directory: '/home/hoo104/LIFT/results/img1_kp.txt'
Parameters
#--------------------------------------------------------------------------------------
Be careful as we do not use sophisticated parsing
The parser will read until the semicolon
types are defined as below
ss: multiple strings separated with commas
s: string
b: boolean
f: float
d: int
#--------------------------------------------------------------------------------------
#--------------------------------------------------------------------------------------
Dataset parameters
ss: dataset.trainSetList = ECCV/piccadilly/; # All the first images of oxford dataset is used for training
s: dataset.dataType = ECCV; # the dataset type
#--------------------------------------------------------------------------------------
Model parameters
s: model.modelType = Combined; # the network type
b: model.bNormalizeInput = 1; # Normalize input to have zero mean and 1 std
f: model.fScaleList = np.array([1.0]); # list of scale spaces (small means larger scale)
GHH related
f: model.max_strength = -1; # GHH softmax strength (-1 for hard)
Keypoints
s: model.sDetector = tilde; # network architecture for kp
d: model.nFilterSize = 25; # Let's keep the number odd
d: model.nPatchSizeKp = 48; # patch size for kp including moving
# regions we use something smaller
s: model.sKpNonlinearity = None; # use nonlinearity at end
f: model.fScaleList = np.array([1.0]); # list of scale spaces (small means larger scale)
f: model.bias_rnd = 0.0; # random noise added to bias
f: model.epsilon = 1e-8; # epsilon for tilde
Orientation
s: model.sOrientation = cvpr16; # network architecture for orientation
Descriptor
s: model.sDescriptor = descriptor_imported; # network architecture for desc
d: model.nDescInputSize = 64; # Input size to be fed to the descriptor
s: model.descriptor_export_folder = /cvlabdata1/home/trulls-data/kpdesc/torch/export/;
L-> skipped
s: model.descriptor_model = new-CNN3-picc-iter-56k.h5; # network configuration
s: model.descriptor_input = desc-input; # Descriptor input
#--------------------------------------------------------------------------------------
Parameters for patch extraction
automatically determined
f: patch.fMaxScale = np.max(self.model.fScaleList); # asserts make sure this is stored properly
f: patch.fRatioScale = (self.model.nPatchSizeKp / 2.0) / 2.0; # to not resize when scale is 2
d: patch.nPatchSize = np.round(self.model.nDescInputSize * self.patch.fRatioScale / 6.0); # the large patch size for data. the desc will look at ratio scale of 6.0
#--------------------------------------------------------------------------------------
Validation and test time parameters
d: validation.batch_size = 100; # batch size of the implementation
d: validation.nScaleInterval = 4;
L-> skipped
d: validation.nNMSInterval = 2; # number of intervals we look for
L-> skipped
# NMS (3 would mean it goes up one
# octave in case of
# nScaleInterval=2)
#---------------------------------------------------------------------------------
Test Data Module
Traceback (most recent call last):
File "compute_descriptor.py", line 111, in
test_data_in = data_module.data_obj(param, image_file_name, kp_file_name)
File "/home/hoo104/LIFT/python-code/Utils/dataset_tools/test.py", line 64, in init
self.load_data(param, image_file_name, kp_file_name)
File "/home/hoo104/LIFT/python-code/Utils/dataset_tools/test.py", line 82, in load_data
pathconf, param, image_file_name, kp_file_name)
File "/home/hoo104/LIFT/python-code/Utils/dataset_tools/test.py", line 129, in load_data_for_set
kp = np.asarray(loadKpListFromTxt(kp_file_name))
File "/home/hoo104/LIFT/python-code/Utils/kp_tools.py", line 196, in loadKpListFromTxt
kp_file = open(kp_file_name, 'rb')
FileNotFoundError: [Errno 2] No such file or directory: '/home/hoo104/LIFT/results/img1_ori.txt'
@yiran-THU
@djidan10
@navia1991
Hello, I think maybe you can help me to solve this problem. Thank you very much.
I have solved the problems.
Refer to : #4
I know the software need work on python2. So, I install anaconda2, and solve it.
Hi,
You would need to do some matching scheme with the keypoints and descriptors. Please see the benchmark_orientations repo for example on how to use it.
Thanks for your reply.
@kmyid
@Hoototo have you try to use benchmark_orientations to realize the match?
@kmyid I get the img_kp.txt, img1_ori.txt and img1_desc.h5. but i do not know the direct relationship between these documents. and the img_kp_txt_scores.h5 file shows -Infinity. thanks
Hi,
@YuYuYan
I want to use benchmark_orientations to realize the match. But I am not have enough time to do this work now.
@Hoototo Thanks for your reply. And I want to realize the match,but I do not know if I can succeed.
@YuYuYan
I think you can do it.
kmyid say that you can see the benchmark_orientations repo for example on how to use it.
https://github.com/cvlab-epfl/learn-orientation
https://github.com/cvlab-epfl/benchmark-orientation
ok,and now I will try to do it,good luck to me , thanks again.
@YuYuYan @Hoototo
Just a quick search on google. All you need is reading .h5 with hdf5read() in matlab (I believe there is other package for Python as well).
keyPt = hdf5read('img1_desc.h5','/keypoints');
desc = hdf5read('img1_desc.h5','/descriptors');
The first two rows in keyPt are x and y coordinates.
Hope this is helpful.
@Hoototo @kmyid @NEU-Gou hi , I need your help
now I try to run the benchmark_orientations ,and I run run_evaluate.m ,but there was an error----- from parse import parse . ImportError: No module named parse.
my python 2.7.13 and I try to import parse in python,it worked. but the error still
Thanks,thanks,thanks
"If we have an error from python execution, retry five times and spit error as the error might just be from theano compile issues... " here is the error
Hello, please check the requirements session. You are missing parse module for python.
But I have import the parse with python,and it succeed
Probably you have some environment issues with virtual environments and such then... cannot help you more about that as it is related to your setup.
Ok I will try it again.Thank you .Thank you for reply.你非常棒!佩服你!
@NEU-Gou Thanks for your reply . Now I want to know use this:
keyPt = hdf5read('img1_desc.h5','/keypoints');
The first two rows in keyPt are x and y coordinates. But what the other rows means?
Hi, @NEU-Gou
I use your method to read .h5 file. In keyPt, the other rows that exclude the first two rows means what?
@Hoototo @YuYuYan I guess those are related to the scale/rotation of the key points.
@kmyid If you guys can provide a little bit more details, that will be really appreciated.
Hi guys, you can have a look at kp_tools.py for details on this. sorry for the late reply.
I have solved the problems.
Refer to : #4
I know the software need work on python2. So, I install anaconda2, and solve it.
学长您好,不知道是否毕业,可以请教一下最后怎么解决的吗
你好,请问有什么问题?
Sorry, Sir. thanks for your reminding.
你好,请问有什么问题?
我也遇到了您提问的问题,不知道怎么解决,求指点 IOError: [Errno 2] No such file or directory '/home/zhuoli/LIFT-master/results/img1_ori.txt' ,方便加我QQ吗,3332146855,感激不尽
@zhuoligetu123
Hi, I think you need to install python2 and have a try.