asanakoy / deeppose_tf Goto Github PK

Hello,

I was wondering if I could use my own images for prediction. I tried to run the test_snapshot.py. I'm looking for a result similar to #2 where the images are visualized. Could someone help me out?

Thank you.

How much iteration do you recommand for training mpii daataset?

Hello,

When I run the file examples/train_mpii_alexnet_imagenet.py, it returns a syntax error at line 43.

Would be great if someone could provide some help!

Thank you.

--2017-10-14 17:34:56-- https://engineering.leeds.ac.uk/info/20132/school_of_computing
Resolving engineering.leeds.ac.uk (engineering.leeds.ac.uk)... 129.11.26.47
Connecting to engineering.leeds.ac.uk (engineering.leeds.ac.uk)|129.11.26.47|:443... connected.
HTTP request sent, awaiting response... 200 OK
Length: unspecified [text/html]
Saving to: ‘lsp_dataset_original.zip’

lsp_dataset_original.zip [ <=> ] 76.47K 19.8KB/s in 3.9s

2017-10-14 17:35:08 (19.8 KB/s) - ‘lsp_dataset_original.zip’ saved [78308]

Archive: lsp_dataset_original.zip
End-of-central-directory signature not found. Either this file is not
a zipfile, or it constitutes one disk of a multi-part archive. In the
latter case the central directory and zipfile comment will be found on
the last disk(s) of this archive.
unzip: cannot find zipfile directory in one of lsp_dataset_original.zip or
lsp_dataset_original.zip.zip, and cannot find lsp_dataset_original.zip.ZIP, period.
mkdir: cannot create directory ‘lsp’: File exists
mv: cannot stat 'images': No such file or directory
mv: cannot stat 'joints.mat': No such file or directory
mv: cannot stat 'README.txt': No such file or directory
--2017-10-14 17:35:09-- http://www.comp.leeds.ac.uk/mat4saj/lspet_dataset.zip
Resolving www.comp.leeds.ac.uk (www.comp.leeds.ac.uk)... 129.11.133.104
Connecting to www.comp.leeds.ac.uk (www.comp.leeds.ac.uk)|129.11.133.104|:80... connected.
HTTP request sent, awaiting response... 302 Found
Location: https://engineering.leeds.ac.uk/info/20132/school_of_computing [following]
--2017-10-14 17:35:09-- https://engineering.leeds.ac.uk/info/20132/school_of_computing
Resolving engineering.leeds.ac.uk (engineering.leeds.ac.uk)... 129.11.26.47
Connecting to engineering.leeds.ac.uk (engineering.leeds.ac.uk)|129.11.26.47|:443... connected.
HTTP request sent, awaiting response... 200 OK
Length: unspecified [text/html]
Saving to: ‘lspet_dataset.zip’

lspet_dataset.zip [ <=> ] 76.47K 169KB/s in 0.5s

2017-10-14 17:35:11 (169 KB/s) - ‘lspet_dataset.zip’ saved [78308]

Archive: lspet_dataset.zip
End-of-central-directory signature not found. Either this file is not
a zipfile, or it constitutes one disk of a multi-part archive. In the
latter case the central directory and zipfile comment will be found on
the last disk(s) of this archive.
unzip: cannot find zipfile directory in one of lspet_dataset.zip or
lspet_dataset.zip.zip, and cannot find lspet_dataset.zip.ZIP, period.
mkdir: cannot create directory ‘lsp_ext’: File exists
mv: cannot stat 'images': No such file or directory
mv: cannot stat 'joints.mat': No such file or directory
mv: cannot stat 'README.txt': No such file or directory

Anyone could share the bvlc_alexnet.tf file to me?
Thank you in advance.

In scripts/train.py

....
    print 'args.resume: {}\nargs.snapshot: {}'.format(args.resume, args.snapshot)
    bbox_extension_range = (args.bbox_extension_min, args.bbox_extension_max)
    if bbox_extension_range[0] is None or bbox_extension_range[1] is None:
        bbox_extension_range = None
        test_bbox_extension_range = None
    else:
        test_bbox_extension_range = (bbox_extension_range[1], bbox_extension_range[1])
...

Why do you use bbox_extension range[1] for the last line, is it a typo?
It should be?

...
 else:
        test_bbox_extension_range = (bbox_extension_range[0], bbox_extension_range[1])
...

Thanks.

I tried to instal Cupy with python 2.7 on windows 10 but i am getting following error

C:\Users\acer.LAPTOP-DAK58UAQ>pip2 install cupy
Collecting cupy
Downloading cupy-2.5.0.tar.gz (1.8MB)
100% |################################| 1.8MB 161kB/s
Complete output from command python setup.py egg_info:
**************************************************
*** WARNING: Cannot find nvToolsExt. nvtx was disabled.
**************************************************
Options: {'profile': False, 'annotate': False, 'linetrace': False, 'no_cuda': False}
Include directories: ['C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v8.0\include', 'C:\Program Files\NVIDIA Corporation\NvToolsExt\include']
Library directories: ['C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v8.0\bin', 'C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v8.0\lib\x64', 'C:\Program Files\NVIDIA Corporation\NvToolsExt\lib\x64']
Microsoft Visual C++ 9.0 is required. Get it from http://aka.ms/vcpython27
**************************************************
*** WARNING: Include files not found: ['cublas_v2.h', 'cuda.h', 'cuda_profiler_api.h', 'cuda_runtime.h', 'curand.h', 'cusparse.h', 'nvrtc.h']
*** WARNING: Skip installing cuda support
*** WARNING: Check your CFLAGS environment variable
**************************************************
Traceback (most recent call last):
File "", line 1, in
File "c:\users\acer.laptop-dak58uaq\appdata\local\temp\pip-build-66dhvz\cupy\setup.py", line 32, in
ext_modules = cupy_setup_build.get_ext_modules()
File "cupy_setup_build.py", line 385, in get_ext_modules
extensions = make_extensions(arg_options, compiler, use_cython)
File "cupy_setup_build.py", line 275, in make_extensions
raise Exception('Your CUDA environment is invalid. '
Exception: Your CUDA environment is invalid. Please check above error log.

----------------------------------------

Command "python setup.py egg_info" failed with error code 1 in c:\users\acer.laptop-dak58uaq\appdata\local\temp\pip-build-66dhvz\cupy\

i have used cuda 8 with openpose as well and it works fine for me but in case of deep pose i am getting the above error.

I train the LSP dataset with imagenet initialization. It has been 260k iterations, the train/pose_error is around[0.02 0.03]. The mPCP score (0.42) is still far from the result listed in the table. It does increase, but pretty slowly.

Does it mean it cannot converge for this time? I need restart the training? Thanks.

I setup everything as you mentioned. But I got the following error.

I tried to test trained lsp datasets during training

wonjinlee@alpha:~/deeppose/out/lsp_alexnet_imagenet_small$ ls
checkpoint events.out.tfevents.1510238719.alpha
checkpoint-100000.data-00000-of-00001 params.dump_171108_222950.txt
checkpoint-100000.index params.dump_171108_223930.txt
checkpoint-100000.meta params.dump_171108_224108.txt
checkpoint-110000.data-00000-of-00001 params.dump_171108_224641.txt
checkpoint-110000.index params.dump_171109_002231.txt
checkpoint-110000.meta params.dump_171109_020558.txt
checkpoint-120000.data-00000-of-00001 params.dump_171109_034216.txt
checkpoint-120000.index params.dump_171109_043955.txt
checkpoint-120000.meta params.dump_171109_060922.txt
checkpoint-130000.data-00000-of-00001 params.dump_171109_061701.txt
checkpoint-130000.index params.dump_171109_145127.txt
checkpoint-130000.meta params.dump_171109_145344.txt
checkpoint-90000.data-00000-of-00001 params.dump_171109_145635.txt
checkpoint-90000.index params.dump_171109_170839.txt
checkpoint-90000.meta params.dump_171109_234514.txt

But it shows this kind of error.

2017-11-10 17:42:08.970095: W tensorflow/core/framework/op_kernel.cc:1192] Data loss: Unable to open table file out/lsp_alexnet_imagenet_small/checkpoint: Data loss: not an sstable (bad magic number): perhaps your file is in a different file format and you need to use a different restore operator?
Traceback (most recent call last):
File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/client/session.py", line 1327, in _do_call
return fn(*args)
File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/client/session.py", line 1306, in _run_fn
status, run_metadata)
File "/usr/lib/python3.5/contextlib.py", line 66, in exit
next(self.gen)
File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/framework/errors_impl.py", line 466, in raise_exception_on_not_ok_status
pywrap_tensorflow.TF_GetCode(status))
tensorflow.python.framework.errors_impl.DataLossError: Unable to open table file out/lsp_alexnet_imagenet_small/checkpoint: Data loss: not an sstable (bad magic number): perhaps your file is in a different file format and you need to use a different restore operator?
[[Node: save/RestoreV2_5 = RestoreV2[dtypes=[DT_FLOAT], _device="/job:localhost/replica:0/task:0/cpu:0"](_arg_save/Const_0_0, save/RestoreV2_5/tensor_names, save/RestoreV2_5/shape_and_slices)]]
[[Node: save/RestoreV2/_37 = _Recvclient_terminated=false, recv_device="/job:localhost/replica:0/task:0/gpu:0", send_device="/job:localhost/replica:0/task:0/cpu:0", send_device_incarnation=1, tensor_name="edge_74_save/RestoreV2", tensor_type=DT_FLOAT, _device="/job:localhost/replica:0/task:0/gpu:0"]]

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "tests/test_snapshot.py", line 116, in
main(dataset_name, snapshot_path)
File "tests/test_snapshot.py", line 79, in main
test_net(test_dataset, test_iterator, dataset_name, snapshot_path)
File "tests/test_snapshot.py", line 92, in test_net
gpu_memory_fraction=0.32) # Set how much GPU memory to reserve for the network
File "/home/wonjinlee/deeppose/scripts/regressionnet.py", line 94, in create_regression_net
saver.restore(net.sess, init_snapshot_path)
File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/training/saver.py", line 1560, in restore
{self.saver_def.filename_tensor_name: save_path})
File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/client/session.py", line 895, in run
run_metadata_ptr)
File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/client/session.py", line 1124, in _run
feed_dict_tensor, options, run_metadata)
File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/client/session.py", line 1321, in _do_run
options, run_metadata)
File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/client/session.py", line 1340, in _do_call
raise type(e)(node_def, op, message)
tensorflow.python.framework.errors_impl.DataLossError: Unable to open table file out/lsp_alexnet_imagenet_small/checkpoint: Data loss: not an sstable (bad magic number): perhaps your file is in a different file format and you need to use a different restore operator?
[[Node: save/RestoreV2_5 = RestoreV2[dtypes=[DT_FLOAT], _device="/job:localhost/replica:0/task:0/cpu:0"](_arg_save/Const_0_0, save/RestoreV2_5/tensor_names, save/RestoreV2_5/shape_and_slices)]]
[[Node: save/RestoreV2/_37 = _Recvclient_terminated=false, recv_device="/job:localhost/replica:0/task:0/gpu:0", send_device="/job:localhost/replica:0/task:0/cpu:0", send_device_incarnation=1, tensor_name="edge_74_save/RestoreV2", tensor_type=DT_FLOAT, _device="/job:localhost/replica:0/task:0/gpu:0"]]

Caused by op 'save/RestoreV2_5', defined at:
File "tests/test_snapshot.py", line 116, in
main(dataset_name, snapshot_path)
File "tests/test_snapshot.py", line 79, in main
test_net(test_dataset, test_iterator, dataset_name, snapshot_path)
File "tests/test_snapshot.py", line 92, in test_net
gpu_memory_fraction=0.32) # Set how much GPU memory to reserve for the network
File "/home/wonjinlee/deeppose/scripts/regressionnet.py", line 93, in create_regression_net
saver = tf.train.Saver()
File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/training/saver.py", line 1140, in init
self.build()
File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/training/saver.py", line 1172, in build
filename=self._filename)
File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/training/saver.py", line 688, in build
restore_sequentially, reshape)
File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/training/saver.py", line 407, in _AddRestoreOps
tensors = self.restore_op(filename_tensor, saveable, preferred_shard)
File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/training/saver.py", line 247, in restore_op
[spec.tensor.dtype])[0])
File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/ops/gen_io_ops.py", line 663, in restore_v2
dtypes=dtypes, name=name)
File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/framework/op_def_library.py", line 767, in apply_op
op_def=op_def)
File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/framework/ops.py", line 2630, in create_op
original_op=self._default_original_op, op_def=op_def)
File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/framework/ops.py", line 1204, in init
self._traceback = self._graph._extract_stack() # pylint: disable=protected-access

DataLossError (see above for traceback): Unable to open table file out/lsp_alexnet_imagenet_small/checkpoint: Data loss: not an sstable (bad magic number): perhaps your file is in a different file format and you need to use a different restore operator?
[[Node: save/RestoreV2_5 = RestoreV2[dtypes=[DT_FLOAT], _device="/job:localhost/replica:0/task:0/cpu:0"](_arg_save/Const_0_0, save/RestoreV2_5/tensor_names, save/RestoreV2_5/shape_and_slices)]]
[[Node: save/RestoreV2/_37 = _Recvclient_terminated=false, recv_device="/job:localhost/replica:0/task:0/gpu:0", send_device="/job:localhost/replica:0/task:0/cpu:0", send_device_incarnation=1, tensor_name="edge_74_save/RestoreV2", tensor_type=DT_FLOAT, _device="/job:localhost/replica:0/task:0/gpu:0"]]

Why this kind of error happens?
Test doesn't work while training?
How can I resolve this error?

Hello, I noticed that you marked that the code is for the first stage.
However, there are three stages in total in the original paper, how can I realize the next two stages by your code?? Maybe just repeat the code by three times??
Looking forward to your reply! Thank you~

Hi, I use the train_lsp_alexnet_imagenet.sh to train. But it seems the training error get stucks to [0.1 0.11] and cannot decrease any more.

If i have to use this code and test a non mpii data set which does not have any attributes ( the position of the various body parts), can it be done? Can i make some changes and feed in images and then see if the different parts of the body are identified.

I am unable to find TensorFlow version 1.0 for python 2.7. Please help how shall I proceed next?

Is it possible to use it for predicting an action from a real-time video feed?(something like checking if a person is running)
I am very new to this field and even if can give me an idea, it would be great.
Thanks

I try to run the implementation with python 3.5 on windows 10.
I have to change some function to change from python 2.x to python 3.5 such as:
print "hello" ---> print("hello")
xrange() --> range()
izip --> zip
Change some import style
Finally, I can run data preparation.
Then when I try to train network from lsp dataset and pretrained AlexNet.
+It load the input data successfully.
+Then, the error come when it run the code in function evaluate_pcp in regressionnet.py.
for i, batch in tqdm(enumerate(test_it), total=num_batches):
I think it comes from the reading data with multithread, but I don't know where the bug come from, and how can I debug the error.
Please guide me how can I check the error, thank you.

The following is the stack trace:

Reading dataset from datasets/lsp_ext/train_joints.csv
8046it [00:43, 184.71it/s]G:\Pose\deeppose_tf-master\scripts\dataset.py:111: Use
rWarning: Skipping joint with incorrect joints coordinates. They are out of the
image.
image: G:/Pose/deeppose_tf-master\datasets/lsp_ext\images\im08075.jpg, joint: [
386.32211538  150.625     ], im.shape: (161, 241)
  'image: {}, joint: {}, im.shape: {}'.format(img_path, joints[i_joint], image_s
hape[:2]))
11000it [00:59, 186.13it/s]
Joints shape: (14, 2)
Reading dataset from datasets/lsp_ext/test_joints.csv
1000it [00:11, 83.85it/s]
Joints shape: (14, 2)
Reading dataset from datasets/lsp_ext/train_lsp_small_joints.csv
1000it [00:10, 99.88it/s]
Joints shape: (14, 2)
1000
<enumerate object at 0x000000002BE6D168>
  0%|                                                    | 0/8 [00:00<?, ?it/s]
Traceback (most recent call last):
File "scripts/train.py", line 242, in <module>
    main(sys.argv[1:])
  File "scripts/train.py", line 237, in main
    output_dir=args.o_dir
  File "scripts/train.py", line 77, in train_loop
    tag_prefix='test')
  File "G:\Pose\deeppose_tf-master\scripts\regressionnet.py", line 279, in evalu
ate_pcp
    for i, batch in tqdm(enumerate(test_it), total=num_batches):
  File "C:\Users\icomlab\AppData\Local\Programs\Python\Python35\lib\site-package
s\tqdm\_tqdm.py", line 959, in __iter__
    for obj in iterable:
  File "C:\Users\icomlab\AppData\Local\Programs\Python\Python35\lib\site-package
s\chainer\iterators\multiprocess_iterator.py", line 87, in __next__
    self._thread = self._prefetch_loop.launch_thread()
  File "C:\Users\icomlab\AppData\Local\Programs\Python\Python35\lib\site-package
s\chainer\iterators\multiprocess_iterator.py", line 307, in launch_thread
    initargs=(self.dataset, self.mem_size, self.mem_bulk))
  File "C:\Users\icomlab\AppData\Local\Programs\Python\Python35\lib\multiprocess
ing\context.py", line 118, in Pool
    context=self.get_context())
  File "C:\Users\icomlab\AppData\Local\Programs\Python\Python35\lib\multiprocess
ing\pool.py", line 168, in __init__
    self._repopulate_pool()
  File "C:\Users\icomlab\AppData\Local\Programs\Python\Python35\lib\multiprocess
ing\pool.py", line 233, in _repopulate_pool
    w.start()
  File "C:\Users\icomlab\AppData\Local\Programs\Python\Python35\lib\multiprocess
ing\process.py", line 105, in start
    self._popen = self._Popen(self)
  File "C:\Users\icomlab\AppData\Local\Programs\Python\Python35\lib\multiprocess
ing\context.py", line 313, in _Popen
    return Popen(process_obj)
  File "C:\Users\icomlab\AppData\Local\Programs\Python\Python35\lib\multiprocess
ing\popen_spawn_win32.py", line 66, in __init__
    reduction.dump(process_obj, to_child)
  File "C:\Users\icomlab\AppData\Local\Programs\Python\Python35\lib\multiprocess
ing\reduction.py", line 59, in dump
    ForkingPickler(file, protocol).dump(obj)
# OSError: [Errno 22] Invalid argument

Traceback (most recent call last):
  File "<string>", line 1, in <module>
  File "C:\Users\icomlab\AppData\Local\Programs\Python\Python35\lib\multiprocess
ing\spawn.py", line 106, in spawn_main
    exitcode = _main(fd)
  File "C:\Users\icomlab\AppData\Local\Programs\Python\Python35\lib\multiprocess
ing\spawn.py", line 116, in _main
    self = pickle.load(from_parent)
# EOFError: Ran out of input

#2
Such as different line color with different joints

the maii dataset the official dataset dont split it as train or test????
I have to split it myself??
different people will have different train/test data????

  bash examples/train_lsp_alexnet_imagenet.sh 

Elapsed time for finding uninitialized variables: 0.42s
Elapsed time to init them: 0.16s
args.resume: False
args.snapshot: /home/wonjinlee/deeppose/weights/bvlc_alexnet.tf
Reading dataset from /home/wonjinlee/data/lsp/train_joints.csv
0it [00:00, ?it/s]
Traceback (most recent call last):
File "/home/wonjinlee/deeppose/scripts/train.py", line 233, in
main(sys.argv[1:])
File "/home/wonjinlee/deeppose/scripts/train.py", line 168, in main
downscale_height=args.downscale_height
File "/home/wonjinlee/deeppose/scripts/dataset.py", line 37, in init
self.load_images()
File "/home/wonjinlee/deeppose/scripts/dataset.py", line 129, in load_images
print('Joints shape:', self.joints[0][1].shape)
IndexError: list index out of range

How could I deal with this kind of error?

the link https://hcicloud.iwr.uni-heidelberg.de/index.php/s/sNklPpCiqOYOCAz cannot be accessed. Anyone could share the bvlc_alexnet.tf file to me? Thank you in advance.

I am trying to create a system that detects only specific parts of human-like bodies; arms to be precise. Is it possible to train this via images of subject specified above?

after running the command python datasets/mpii_dataset.py I am getting the error as no module named scripts.config.
Here is what I got when i run the command.
aniket@LAPTOP-DAK58UAQ:~/deeppose_tf$ python datasets/mpii_dataset.py
Traceback (most recent call last):
File "datasets/mpii_dataset.py", line 14, in
from scripts.config import *
ImportError: No module named scripts.config

Hi after training on LSP dataset.
When I output the prediction for different image, the output joint location are really close.
The reason could be I just train for 20000 iteration.
Could you let me know the reasonable training iteration or could you provide trained weights?
Thx