guanghan/lighttrack

LightTrack: A Generic Framework for Online Top-Down Human Pose Tracking

lighttrack's Issues

Dose this support MacOS CPU version and ignore the inference speed?

I want to test this demo on my MacOS, however when i "make" in "lib" folder it shows that it requires cuda nvcc home.

how to prepare pickle data for train Pose Matching Module

To train Pose Matching Module, we need two pickle data, which are

 data_path: ./unit_test/posetrack_train_data.pickle
 data_neg_path: ./unit_test/posetrack_train_data_negative.pickle

I think test_keypoints_to_graph.py can make the pickle, but it need to read json files from 'data/Data_2018/posetrack_data/gcn_openSVAI/train/*.json'. My question is, where these json files come from? Am I right? Look forward to your reply.

Any recommendation to make it on Keras?

Hi,

I am trying to make it work on Keras. Are there any recommendation? Or it will never work?

Bests,
Than

How to merge annotation json file？

Hi,@Guanghan thanks for your great work!
I have a problem, as the downloaded annotation files are separated json files, which way do you merge them into one file? Could you give the code to merge these files?
Thanks very much!

How to run the camera inference in online mode?

Hi Guanghan,

Thanks for your work. Now in demo_video_mobile.py, we read an video and process with online mode (one by one frame.)

Do you have scripts with run the model with a camera? You will get one frame each time without knowing the total number of frames.

Thank you.

download error 403

I meet 403 error when trying to download the pretrained model in your website. Is the link still valid?

str2bool / DictAction (from torchlight)

I assume for your demo script, there is an error with the torchlight and the IO functions from:
https://github.com/yysijie/st-gcn/blob/master/torchlight/torchlight/io.py

I get this error:

lighttrack/graph/gcn_utils/io.py", line 18, in <module>
    from torchlight import str2bool
ImportError: cannot import name 'str2bool'

But those "to be imported" modules are found there:
https://github.com/yysijie/st-gcn/blob/master/torchlight/torchlight/io.py

IndexError: list index out of range

I just perform LightTrack demo on my own video, and yolov3 detects nothing in one frame, then this problem appear.

Where is evaluate.py for evaluation on PoseTrack 2018?

Thanks for your great work.
I want to evaluation on PoseTrack 2018, but where is evaluate.py and "data/Data_2018/poseval/py "?

Issue while Running this command '!python /content/lighttrack/demo_video_mobile.py'

I Installed all required file but still having issue..

File "/content/lighttrack/demo_video_mobile.py", line 32, in
from graph import visualize_pose_matching
File "/content/lighttrack/graph/visualize_pose_matching.py", line 22, in
from utils_json import *
ModuleNotFoundError: No module named 'utils_json'

About demo_video_mobile.py running

Hello, I am interested in your work, so I want to run demo_video_mobile.py.Flow your setps, I downloads all files, I meet a problem when I run demo_video_mobile.py. It is " No such file or directory: './graph/GCN/epoch210_model.pt'", I don't know how to deal with it, please help me, I'll be very grateful with it. Another, I run in windows.Thanks!

will you release the training code for posetrack18?

if i want to reproduce your result on posetrack18, is there anything i should notice about?

Replace pose estimator with another CPU-based only

Hi,

I would like to replace the pose estimator with a lightweight implementation of OpenPose, based only on CPU inference since my computer does not support CUDA.

However when I try to compile the 'lib' folder with 'make' I get an error due to the missing CUDA support. Is it possible to do the replacement explained above or CUDA is a prerequisite for other parts of Lighttrack too?

Thank you,
Cataldo

Error in running with flag_nms= True

When I run demo_video_mobile.py after setting flag_nms = True I get the following error :-
TypeError: list indices must be integers or slices, not tuple

For the following code inside the function def apply_nms :-
keep = np.where((cls_dets[:, 4] >= min_scores) &
((cls_dets[:, 3] - cls_dets[:, 1]) * (cls_dets[:, 2] - cls_dets[:, 0]) >= min_box_size))[0]

When I first convert cls_dets into numpy array by cls_dets = np.array(cls_dets) and then execute, the following error is shown :-
IndexError: too many indices for array

This is because the shape of cls_dets is (4,) while the code written is for a 2 dimensional array.

Please suggest a solution so that I can apply non max suppresion with this code.

Download link expired

The two links are expired, would you please update the links?

Training

1. Pose Estimation Module
   To train, grab an imagenet-pretrained model and put it in ./weights.
   For Resnet101, download resnet101.ckpt from here.
   For Resnet152, download resnet152.ckpt from here.

Detection result of PoseTrack17 validation set

Can you please provide the detection result on PoseTrack17 validation set using deformable FPN? Thank you very much in advance.

在运行demo_video_mobile.py时报错

(py36) liuzhuxian@liuzhuxian-Inspiron-5559:~/lighttrack$ python demo_video_mobile.py
/home/liuzhuxian/anaconda3/envs/py36/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:516: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint8 = np.dtype([("qint8", np.int8, 1)])
/home/liuzhuxian/anaconda3/envs/py36/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:517: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_quint8 = np.dtype([("quint8", np.uint8, 1)])
/home/liuzhuxian/anaconda3/envs/py36/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:518: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint16 = np.dtype([("qint16", np.int16, 1)])
/home/liuzhuxian/anaconda3/envs/py36/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:519: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_quint16 = np.dtype([("quint16", np.uint16, 1)])
/home/liuzhuxian/anaconda3/envs/py36/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:520: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint32 = np.dtype([("qint32", np.int32, 1)])
/home/liuzhuxian/anaconda3/envs/py36/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:525: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
np_resource = np.dtype([("resource", np.ubyte, 1)])
/home/liuzhuxian/anaconda3/envs/py36/lib/python3.7/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:541: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint8 = np.dtype([("qint8", np.int8, 1)])
/home/liuzhuxian/anaconda3/envs/py36/lib/python3.7/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:542: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_quint8 = np.dtype([("quint8", np.uint8, 1)])
/home/liuzhuxian/anaconda3/envs/py36/lib/python3.7/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:543: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint16 = np.dtype([("qint16", np.int16, 1)])
/home/liuzhuxian/anaconda3/envs/py36/lib/python3.7/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:544: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_quint16 = np.dtype([("quint16", np.uint16, 1)])
/home/liuzhuxian/anaconda3/envs/py36/lib/python3.7/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:545: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint32 = np.dtype([("qint32", np.int32, 1)])
/home/liuzhuxian/anaconda3/envs/py36/lib/python3.7/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:550: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
np_resource = np.dtype([("resource", np.ubyte, 1)])
WARNING:tensorflow:From /home/liuzhuxian/lighttrack/HPE/../lib/nets/mobilenet_v1.py:440: The name tf.GraphKeys is deprecated. Please use tf.compat.v1.GraphKeys instead.

OMP: Info #212: KMP_AFFINITY: decoding x2APIC ids.
OMP: Info #210: KMP_AFFINITY: Affinity capable, using global cpuid leaf 11 info
OMP: Info #154: KMP_AFFINITY: Initial OS proc set respected: 0-3
OMP: Info #156: KMP_AFFINITY: 4 available OS procs
OMP: Info #157: KMP_AFFINITY: Uniform topology
OMP: Info #179: KMP_AFFINITY: 1 packages x 2 cores/pkg x 2 threads/core (2 total cores)
OMP: Info #214: KMP_AFFINITY: OS proc to physical thread map:
OMP: Info #171: KMP_AFFINITY: OS proc 0 maps to package 0 core 0 thread 0
OMP: Info #171: KMP_AFFINITY: OS proc 2 maps to package 0 core 0 thread 1
OMP: Info #171: KMP_AFFINITY: OS proc 1 maps to package 0 core 1 thread 0
OMP: Info #171: KMP_AFFINITY: OS proc 3 maps to package 0 core 1 thread 1
OMP: Info #250: KMP_AFFINITY: pid 2591 tid 2591 thread 0 bound to OS proc set 0
Detector YOLOv3 options: Namespace(batch_size=1, checkpoint_model=None, conf_thres=0.8, config_path='detector/config/yolov3.cfg', img_size=416, n_cpu=8, nms_thres=0.4, weights_path='weights/YOLOv3/yolov3.weights')
OMP: Info #250: KMP_AFFINITY: pid 2591 tid 2604 thread 1 bound to OS proc set 1
Traceback (most recent call last):
File "demo_video_mobile.py", line 28, in
from nms.gpu_nms import gpu_nms
ModuleNotFoundError: No module named 'nms.gpu_nms'

undefined symbol: __cudaRegisterFatBinaryEnd

Hi @Guanghan

I got an error when the python demo_video_mobile.py is implemented.

Traceback (most recent call last):
File "demo_video_mobile.py", line 28, in
from nms.gpu_nms import gpu_nms
ImportError: /home/lighttrack/HPE/../lib/nms/gpu_nms.cpython-36m-x86_64-linux-gnu.so: undefined symbol: __cudaRegisterFatBinaryEnd

how can this issue be solved?

Script does not work if predictions_lighttrack folder does not exist

I can't remember which script it is, but an empty predictions_lighttrack must exist in data/Data_2018 for it to work. I can check again next week which one it is.

Ran out of input

i met a below problem, when i train the pose match. it seems that there are something wrong with the file--"posetrack_train_data_negative.pickle".

 In [3]: with open("posetrack_train_data_negative.pickle", "rb") as f: 
   ...:     a = pickle.load(f) 
   ...:      
   ...:                                                                           
---------------------------------------------------------------------------
EOFError                                  Traceback (most recent call last)
<ipython-input-3-853959a245df> in <module>
      1 with open("posetrack_train_data_negative.pickle", "rb") as f:
----> 2     a = pickle.load(f)
      3 
      4 

EOFError: Ran out of input

while running demo_video_mobile.py, I came across this problem.

Detector YOLOv3 options: Namespace(batch_size=1, checkpoint_model=None, conf_thres=0.8, config_path='detector/config/yolov3.cfg', img_size=416, n_cpu=8, nms_thres=0.4, weights_path='weights/YOLOv3/yolov3.weights')
PoseTrack order.
/home/vdai/peng/code/lighttrack-master/graph/gcn_utils/io.py:43: YAMLLoadWarning: calling yaml.load() without Loader=... is deprecated, as the default Loader is unsafe. Please read https://msg.pyyaml.org/load for full details.
default_arg = yaml.load(f)
[08.14.19|11:47:23] Load weights from ./weights/GCN/epoch210_model.pt.
[08.14.19|11:47:23] Load weights [A].
[08.14.19|11:47:23] Load weights [data_bn.weight].
[08.14.19|11:47:23] Load weights [data_bn.bias].
[08.14.19|11:47:23] Load weights [data_bn.running_mean].
[08.14.19|11:47:23] Load weights [data_bn.running_var].
[08.14.19|11:47:23] Load weights [data_bn.num_batches_tracked].
[08.14.19|11:47:23] Load weights [st_gcn_networks.0.gcn.conv.weight].
[08.14.19|11:47:23] Load weights [st_gcn_networks.0.gcn.conv.bias].
[08.14.19|11:47:23] Load weights [st_gcn_networks.1.gcn.conv.weight].
[08.14.19|11:47:23] Load weights [st_gcn_networks.1.gcn.conv.bias].
[08.14.19|11:47:23] Load weights [edge_importance.0].
[08.14.19|11:47:23] Load weights [edge_importance.1].
[08.14.19|11:47:23] Load weights [fcn.weight].
[08.14.19|11:47:23] Load weights [fcn.bias].
Traceback (most recent call last):
File "/usr/local/lib/python3.5/dist-packages/torchlight-1.0-py3.5.egg/torchlight/io.py", line 82, in load_weights
doc = _io._TextIOBase.doc
File "/home/vdai/.local/lib/python3.5/site-packages/torch/nn/modules/module.py", line 721, in load_state_dict
self.class.name, "\n\t".join(error_msgs)))
RuntimeError: Error(s) in loading state_dict for Model:
Unexpected key(s) in state_dict: "data_bn.num_batches_tracked".

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "/home/vdai/peng/code/lighttrack-master/demo_video_mobile.py", line 32, in
from graph import visualize_pose_matching
File "/home/vdai/peng/code/lighttrack-master/graph/visualize_pose_matching.py", line 297, in
pose_matcher = Pose_Matcher()
File "/home/vdai/peng/code/lighttrack-master/graph/visualize_pose_matching.py", line 162, in init
self.load_weights()
File "/home/vdai/peng/code/lighttrack-master/graph/gcn_utils/io.py", line 79, in load_weights
self.arg.ignore_weights)
File "/usr/local/lib/python3.5/dist-packages/torchlight-1.0-py3.5.egg/torchlight/io.py", line 89, in load_weights

File "/home/vdai/.local/lib/python3.5/site-packages/torch/nn/modules/module.py", line 721, in load_state_dict
self.class.name, "\n\t".join(error_msgs)))
RuntimeError: Error(s) in loading state_dict for Model:
Unexpected key(s) in state_dict: "data_bn.num_batches_tracked".

Full path contained in DeformConv_RFCN_detect.zip

http://guanghan.info/download/Data/LightTrack/dets/DeformConv_RFCN_detect.zip contains full paths rather than relative ones.

Validation set size inconsistent with PoseTrack2018

Hi,

Thank you for releasing this quality work!

I notices in your paper you mentioned that validation set of PoseTrack18 has 74 sequences, while the official PoseTrack dataset has 170 validation sequences. I assume all the results regarding the validation set here and in the paper correspond to 74 sequences, not the full 170 sequences? Can you provide a way to evaluate your models (detection, pose estimation, matching, etc.) on full validation set?

thanks!

score>1?

Thanks for openning source the awesome code!
when I run demo_camera_mobile.py , found some keypoints score > 1, code in there https://github.com/Guanghan/lighttrack/blob/master/demo_camera_mobile.py#L619
Is this you set deliberatly?

How can I get the key points in the inference phase?

Thanks for your great job first! In my task, I need to extract the key points in the video, so which files should I read? Or does any utils in this program offer for the reader to extract key points? Thank you very much.

ImportError: gpu_nmscpython-36m-x86_64-linux-gnu.so: undefined symbol: __cudaPopCallConfiguration

Hi,
I am facing the following error when I run demo_video_mobile.py .

Traceback (most recent call last): File "demo_video_mobile.py", line 28, in <module> from nms.gpu_nms import gpu_nms ImportError: /home/peeterson/Desktop/lighttrack/HPE/../lib/nms/gpu_nms.cpython-36m-x86_64-linux-gnu.so: undefined symbol: __cudaPopCallConfiguration

module versions inside py36 environment:

_libgcc_mutex 0.1 conda_forge conda-forge
_openmp_mutex 4.5 1_llvm conda-forge
_tflow_180_select 1.0 gpu anaconda
absl-py 0.2.2 py36_0 anaconda
astor 0.6.2 py36_0 anaconda
blas 1.0 mkl conda-forge
bleach 1.5.0 py36_0 conda-forge
bzip2 1.0.6 1 conda-forge
ca-certificates 2020.12.5 ha878542_0 conda-forge
cairo 1.14.12 h77bcde2_0 anaconda
certifi 2020.12.5 py36h5fab9bb_1 conda-forge
cloudpickle 0.5.3 pypi_0 pypi
cmake 3.12.0 h307fef2_0 anaconda
cudatoolkit 9.0 h13b8566_0 anaconda
cudnn 7.1.2 cuda9.0_0 anaconda
cupti 9.0.176 0 anaconda
cycler 0.10.0 py36_0 conda-forge
cython 0.28.3 pypi_0 pypi
dask 0.17.5 pypi_0 pypi
dataclasses 0.8 pypi_0 pypi
datetime 4.2 pypi_0 pypi
dbus 1.13.0 h3a4f0e9_0 conda-forge
decorator 4.3.0 pypi_0 pypi
easydict 1.7 pypi_0 pypi
expat 2.2.5 0 conda-forge
ffmpeg 3.2.4 3 conda-forge
fontconfig 2.13.0 0 conda-forge
freetype 2.8.1 0 conda-forge
gast 0.2.0 py36_0 anaconda
giflib 5.1.4 0 conda-forge
glib 2.53.6 h5d9569c_2 anaconda
graphite2 1.3.11 0 conda-forge
grpcio 1.12.0 py36hdbcaa40_0 anaconda
gst-plugins-base 1.12.4 h33fb286_0 anaconda
gstreamer 1.12.4 hb53b477_0 anaconda
h5py 2.9.0 pypi_0 pypi
harfbuzz 1.7.6 hc5b324e_0 anaconda
hdf5 1.10.2 0 conda-forge
html5lib 0.9999999 py36_0 conda-forge
icu 58.2 0 conda-forge
imageio 2.3.0 py36_0 conda-forge
jasper 1.900.1 4 conda-forge
jpeg 9b h024ee3a_2
kiwisolver 1.0.1 py36_1 conda-forge
libcurl 7.61.0 h1ad7b7a_0 anaconda
libedit 3.1.20170329 h6b74fdf_2 anaconda
libffi 3.2.1 hd88cf55_4 anaconda
libgcc 7.2.0 h69d50b8_2 conda-forge
libgcc-ng 7.2.0 hdf63c60_3 anaconda
libgfortran 3.0.0 1 conda-forge
libgfortran-ng 7.2.0 hdf63c60_3 anaconda
libiconv 1.15 h63c8f33_5 anaconda
libopenblas 0.2.20 h9ac9557_7 anaconda
libpng 1.6.34 0 conda-forge
libprotobuf 3.5.2 h6f1eeef_0 anaconda
libssh2 1.8.0 h9cfc8f7_4 anaconda
libstdcxx-ng 7.2.0 hdf63c60_3 anaconda
libtiff 4.0.9 he6b73bb_1 conda-forge
libuuid 1.0.3 h1bed415_2 anaconda
libwebp 0.5.2 7 conda-forge
libxcb 1.13 0 conda-forge
libxml2 2.9.8 h26e45fe_1 anaconda
llvm-openmp 11.0.1 h4bd325d_0 conda-forge
markdown 2.6.11 py36_0 anaconda
matplotlib 2.2.2 py36h0e671d2_0 anaconda
mkl 2020.4 h726a3e6_304 conda-forge
msgpack 0.5.6 pypi_0 pypi
msgpack-numpy 0.4.3 pypi_0 pypi
munkres 1.0.12 pypi_0 pypi
ncurses 6.1 hf484d3e_0 anaconda
networkx 2.1 pypi_0 pypi
numpy 1.14.3 py36h28100ab_1 anaconda
numpy-base 1.14.3 py36h0ea5e3f_1 anaconda
olefile 0.45.1 py36_0 conda-forge
openblas 0.2.20 8 conda-forge
opencv 3.4.1 py36h6fd60c2_1 anaconda
opencv-python 3.4.1.15 pypi_0 pypi
openssl 1.0.2o 0 conda-forge
pcre 8.41 1 conda-forge
pillow 5.1.0 py36_0 conda-forge
pip 21.0.1 pypi_0 pypi
pixman 0.34.0 2 conda-forge
protobuf 3.5.2 py36hf484d3e_0 anaconda
pyarrow 0.9.0 pypi_0 pypi
pycocotools 2.0.0 pypi_0 pypi
pyparsing 2.2.0 py36_0 conda-forge
pyqt 5.6.0 py36_5 conda-forge
pyqt5 5.12.1 pypi_0 pypi
pyqt5-sip 4.19.15 pypi_0 pypi
python 3.6.5 hc3d631a_2 anaconda
python-dateutil 2.7.3 py_0 conda-forge
python_abi 3.6 1_cp36m conda-forge
pytz 2018.4 py_0 conda-forge
pywavelets 0.5.2 pypi_0 pypi
pyyaml 3.12 pypi_0 pypi
pyzmq 17.0.0 pypi_0 pypi
qt 5.6.2 hd25b39d_14 anaconda
readline 7.0 ha6073c6_4 anaconda
rhash 1.3.6 hb7f436b_0 anaconda
scikit-image 0.14.0 pypi_0 pypi
scipy 1.1.0 pypi_0 pypi
setproctitle 1.1.10 pypi_0 pypi
setuptools 39.2.0 py36_0 conda-forge
shapely 1.6.4.post1 pypi_0 pypi
sip 4.18 py36_1 conda-forge
six 1.11.0 py36h372c433_1 anaconda
sqlite 3.23.1 he433501_0 anaconda
tabulate 0.8.2 pypi_0 pypi
tensorboard 1.8.0 py36hf484d3e_0 anaconda
tensorflow 1.8.0 hb11d968_0 anaconda
tensorflow-base 1.8.0 py36hc1a7637_0 anaconda
tensorflow-gpu 1.8.0 h7b35bdc_0 anaconda
termcolor 1.1.0 py36_1 conda-forge
tk 8.6.7 hc745277_3 anaconda
toolz 0.9.0 pypi_0 pypi
torch 1.0.1 pypi_0 pypi
torchvision 0.2.2 pypi_0 pypi
tornado 5.0.2 py36_0 conda-forge
tqdm 4.19.9 pypi_0 pypi
typing-extensions 3.7.4.3 pypi_0 pypi
webencodings 0.5.1 py_1 conda-forge
werkzeug 0.14.1 py36_0 anaconda
wheel 0.31.1 py36_0 conda-forge
x264 20131218 0 conda-forge
xorg-libxau 1.0.8 3 conda-forge
xorg-libxdmcp 1.1.2 3 conda-forge
xz 5.2.4 h14c3975_4 anaconda
zlib 1.2.11 ha838bed_2 anaconda
zope-interface 4.5.0 pypi_0 pypi

I have installed the same cuda and cudnn versions above in my system.

I have exported the path as well.
export PATH=/usr/local/cuda-9.0/bin export LD_LIBRARY_PATH=/usr/local/cuda-9.0/lib64:/usr/local/cuda-9.0/extras/CUPTI/lib64:/lib/nccl/cuda-9

I also tried to use an earlier version of gcc as per this:
[https://stackoverflow.com/questions/6622454/cuda-incompatible-with-my-gcc-version]

I still get the same error. could someone suggest how to fix this?

thanks.

FileNotFoundError: [Errno 2] No such file or directory: './graph/GCN/epoch210_model.pt'

Hello Guanghan,

First of all, thanks for this amazing repo. I followed the instructions to run the demo and run into this error:
FileNotFoundError: [Errno 2] No such file or directory: './graph/GCN/epoch210_model.pt'
The solution is as simple as changing './graph/GCN/epoch210_model.pt' for './weights/GCN/epoch210_model.pt' in lighttrack/graph/config/inference.yaml

Hope it helps. Cheers,
Dunai

Merged dataset training

Pose track datasets have 15 keypoints while COCO has 17, how to merge the dataset and train Jointly?

Best performance config?

How should I make a proper configuration with "MSRA" pose estimitor and "FPN-101" dector in "offline" mode to achieve the best performance ?

EnvironmentError: The CUDA lib64 path could not be located in /usr/lib64

Error occured during make.

#python3.5 setup.py build_ext --inplace
python setup.py build_ext --inplace
Traceback (most recent call last):
  File "setup.py", line 55, in <module>
    CUDA = locate_cuda()
  File "setup.py", line 52, in locate_cuda
    raise EnvironmentError('The CUDA %s path could not be located in %s' % (k, v))
EnvironmentError: The CUDA lib64 path could not be located in /usr/lib64
Makefile:2: recipe for target 'all' failed
make: *** [all] Error 1

您好，请问有18个关键点的骨骼检测模型吗

我想和st-gcn联用来判断具体行为，请问有openpose格式的18个坐标点的输出吗

AttributeError: 'MSVCCompiler' object has no attribute 'compiler_so'

Hello,
I am trying to execute the 'make' command inside the lib folder and i get the following error:
`Traceback (most recent call last):
File "setup.py", line 151, in
cmdclass={'build_ext': custom_build_ext},

File "C:\Users\Chris\anaconda3\envs\studienarbeit_env\lib\distutils\core.py", line 148, in setup
dist.run_commands()
File "C:\Users\Chris\anaconda3\envs\studienarbeit_env\lib\distutils\dist.py", line 955, in run_commands
self.run_command(cmd)
File "C:\Users\Chris\anaconda3\envs\studienarbeit_env\lib\distutils\dist.py", line 974, in run_command
cmd_obj.run()
File "C:\Users\Chris\anaconda3\envs\studienarbeit_env\lib\site-packages\Cython\Distutils\old_build_ext.py", line 186, in run
_build_ext.build_ext.run(self)
File "C:\Users\Chris\anaconda3\envs\studienarbeit_env\lib\distutils\command\build_ext.py", line 339, in run
self.build_extensions()
File "setup.py", line 106, in build_extensions
customize_compiler_for_nvcc(self.compiler)
File "setup.py", line 78, in customize_compiler_for_nvcc
default_compiler_so = self.compiler_so
AttributeError: 'MSVCCompiler' object has no attribute 'compiler_so'
make: *** [all] Fehler 1`

Can anybody help me with this?

How to get the file under the detections_openSVAI path

How to get the file under the detections_openSVAI path？

MaskRCNN and LightTrack

Can you guys provide information on how to switch the CPN and YOLO models with MaskRCNN?

PoseTrack download

Hi

I can't download the PoseTrack dataset, it seems that the PoseTrack official website is currently inaccessible.
Do you have any other accessible PoseTrack data download links to share?

Looking forward to your reply.
Thank you very much.

Missing imports: logger, pprint, etc.

flake8 testing of https://github.com/Guanghan/lighttrack on Python 3.7.1

$ flake8 . --count --select=E9,F63,F72,F82 --show-source --statistics

./lib/tfflat/data_provider.py:359:21: F821 undefined name 'logger'
                    logger.exception("Cannot batch data. Perhaps they are of inconsistent shape?")
                    ^
./lib/tfflat/data_provider.py:361:29: F821 undefined name 'pprint'
                        s = pprint.pformat([x[k].shape for x in data_holder])
                            ^
./lib/tfflat/data_provider.py:362:25: F821 undefined name 'logger'
                        logger.error("Shape of all arrays to be batched: " + s)
                        ^
./HPE/dataset.py:17:19: F821 undefined name 'mask'
            rle = mask.frPyObjects([seg_ann], label.shape[0], label.shape[1])
                  ^
./HPE/dataset.py:21:13: F821 undefined name 'mask'
        m = mask.decode(rle) * 1
            ^
./HPE/dataset.py:231:23: F821 undefined name 'ori_img'
        seg = get_seg(ori_img.shape[0], ori_img.shape[1], d['segmentation'])
                      ^
./HPE/dataset.py:231:41: F821 undefined name 'ori_img'
        seg = get_seg(ori_img.shape[0], ori_img.shape[1], d['segmentation'])
                                        ^
./utils/standard_classes.py:23:20: F821 undefined name 'python_to_json'
        json_str = python_to_json(self.python_data)
                   ^
./utils/standard_classes.py:27:9: F821 undefined name 'write_json_to_file'
        write_json_to_file(self.python_data, output_json_path)
        ^
./utils/standard_classes.py:65:40: F821 undefined name 'pose_order'
        self.candidate["pose_order"] = pose_order
                                       ^
./utils/utils_convert_heatmap.py:138:44: E999 SyntaxError: invalid syntax
        print "[congrid] dimensions error. " \
                                           ^
./utils/utils_pose.py:197:44: E999 TabError: inconsistent use of tabs and spaces in indentation
	draw_heatmap(heatmap, joint_names[ith_map])
                                           ^
./utils/utils_io_file.py:95:16: F821 undefined name 'as_str'
        return as_str.index("1"), 63 - as_str.rindex("1")
               ^
./utils/utils_io_file.py:95:40: F821 undefined name 'as_str'
        return as_str.index("1"), 63 - as_str.rindex("1")
                                       ^
2     E999 SyntaxError: invalid syntax
12    F821 undefined name 'mask'
14

E901,E999,F821,F822,F823 are the "showstopper" flake8 issues that can halt the runtime with a SyntaxError, NameError, etc. These 5 are different from most other flake8 issues which are merely "style violations" -- useful for readability but they do not effect runtime safety.

• F821: undefined name name
• F822: undefined name name in __all__
• F823: local variable name referenced before assignment
• E901: SyntaxError or IndentationError
• E999: SyntaxError -- failed to compile a file into an Abstract Syntax Tree

Can I use it to calculate the number of visitors with a video?

ImportError: /lib/nms/gpu_nms.so: undefined symbol: _Py_ZeroStruct

Error when running the demo_video_mobile.py

Detector YOLOv3 options: Namespace(batch_size=1, checkpoint_model=None, conf_thres=0.8, config_path='detector/config/yolov3.cfg', img_size=416, n_cpu=8, nms_thres=0.4, weights_path='weights/YOLOv3/yolov3.weights')
Traceback (most recent call last):
  File "demo_video_mobile.py", line 28, in <module>
    from nms.gpu_nms import gpu_nms
ImportError: /home/icarus/oliver/lighttrack/HPE/../lib/nms/gpu_nms.so: undefined symbol: _Py_ZeroStruct

I did run the make in lib, the make output is:

#python3.5 setup.py build_ext --inplace
python setup.py build_ext --inplace
running build_ext
skipping 'utils/bbox.c' Cython extension (up-to-date)
building 'utils.cython_bbox' extension
creating build
creating build/temp.linux-x86_64-2.7
creating build/temp.linux-x86_64-2.7/utils
{'gcc': ['-Wno-cpp', '-Wno-unused-function']}
x86_64-linux-gnu-gcc -pthread -DNDEBUG -g -fwrapv -O2 -Wall -Wstrict-prototypes -fno-strict-aliasing -Wdate-time -D_FORTIFY_SOURCE=2 -g -fstack-protector-strong -Wformat -Werror=format-security -fPIC -I/home/icarus/.local/lib/python2.7/site-packages/numpy/core/include -I/usr/include/python2.7 -c utils/bbox.c -o build/temp.linux-x86_64-2.7/utils/bbox.o -Wno-cpp -Wno-unused-function
x86_64-linux-gnu-gcc -pthread -shared -Wl,-O1 -Wl,-Bsymbolic-functions -Wl,-Bsymbolic-functions -Wl,-z,relro -fno-strict-aliasing -DNDEBUG -g -fwrapv -O2 -Wall -Wstrict-prototypes -Wdate-time -D_FORTIFY_SOURCE=2 -g -fstack-protector-strong -Wformat -Werror=format-security -Wl,-Bsymbolic-functions -Wl,-z,relro -Wdate-time -D_FORTIFY_SOURCE=2 -g -fstack-protector-strong -Wformat -Werror=format-security build/temp.linux-x86_64-2.7/utils/bbox.o -o /home/icarus/oliver/lighttrack/lib/utils/cython_bbox.so
skipping 'utils/nms.c' Cython extension (up-to-date)
building 'utils.cython_nms' extension
{'gcc': ['-Wno-cpp', '-Wno-unused-function']}
x86_64-linux-gnu-gcc -pthread -DNDEBUG -g -fwrapv -O2 -Wall -Wstrict-prototypes -fno-strict-aliasing -Wdate-time -D_FORTIFY_SOURCE=2 -g -fstack-protector-strong -Wformat -Werror=format-security -fPIC -I/home/icarus/.local/lib/python2.7/site-packages/numpy/core/include -I/usr/include/python2.7 -c utils/nms.c -o build/temp.linux-x86_64-2.7/utils/nms.o -Wno-cpp -Wno-unused-function
x86_64-linux-gnu-gcc -pthread -shared -Wl,-O1 -Wl,-Bsymbolic-functions -Wl,-Bsymbolic-functions -Wl,-z,relro -fno-strict-aliasing -DNDEBUG -g -fwrapv -O2 -Wall -Wstrict-prototypes -Wdate-time -D_FORTIFY_SOURCE=2 -g -fstack-protector-strong -Wformat -Werror=format-security -Wl,-Bsymbolic-functions -Wl,-z,relro -Wdate-time -D_FORTIFY_SOURCE=2 -g -fstack-protector-strong -Wformat -Werror=format-security build/temp.linux-x86_64-2.7/utils/nms.o -o /home/icarus/oliver/lighttrack/lib/utils/cython_nms.so
skipping 'nms/cpu_nms.c' Cython extension (up-to-date)
building 'nms.cpu_nms' extension
creating build/temp.linux-x86_64-2.7/nms
{'gcc': ['-Wno-cpp', '-Wno-unused-function']}
x86_64-linux-gnu-gcc -pthread -DNDEBUG -g -fwrapv -O2 -Wall -Wstrict-prototypes -fno-strict-aliasing -Wdate-time -D_FORTIFY_SOURCE=2 -g -fstack-protector-strong -Wformat -Werror=format-security -fPIC -I/home/icarus/.local/lib/python2.7/site-packages/numpy/core/include -I/usr/include/python2.7 -c nms/cpu_nms.c -o build/temp.linux-x86_64-2.7/nms/cpu_nms.o -Wno-cpp -Wno-unused-function
x86_64-linux-gnu-gcc -pthread -shared -Wl,-O1 -Wl,-Bsymbolic-functions -Wl,-Bsymbolic-functions -Wl,-z,relro -fno-strict-aliasing -DNDEBUG -g -fwrapv -O2 -Wall -Wstrict-prototypes -Wdate-time -D_FORTIFY_SOURCE=2 -g -fstack-protector-strong -Wformat -Werror=format-security -Wl,-Bsymbolic-functions -Wl,-z,relro -Wdate-time -D_FORTIFY_SOURCE=2 -g -fstack-protector-strong -Wformat -Werror=format-security build/temp.linux-x86_64-2.7/nms/cpu_nms.o -o /home/icarus/oliver/lighttrack/lib/nms/cpu_nms.so
skipping 'nms/gpu_nms.cpp' Cython extension (up-to-date)
building 'nms.gpu_nms' extension
{'gcc': ['-Wno-unused-function'], 'nvcc': ['-arch=sm_52', '--ptxas-options=-v', '-c', '--compiler-options', "'-fPIC'"]}
/usr/bin/nvcc -I/home/icarus/.local/lib/python2.7/site-packages/numpy/core/include -I/usr/include -I/usr/include/python2.7 -c nms/nms_kernel.cu -o build/temp.linux-x86_64-2.7/nms/nms_kernel.o -arch=sm_52 --ptxas-options=-v -c --compiler-options '-fPIC'
ptxas info    : 0 bytes gmem
ptxas info    : Compiling entry function '_Z10nms_kernelifPKfPy' for 'sm_52'
ptxas info    : Function properties for _Z10nms_kernelifPKfPy
    0 bytes stack frame, 0 bytes spill stores, 0 bytes spill loads
ptxas info    : Used 23 registers, 1280 bytes smem, 344 bytes cmem[0], 4 bytes cmem[2]
{'gcc': ['-Wno-unused-function'], 'nvcc': ['-arch=sm_52', '--ptxas-options=-v', '-c', '--compiler-options', "'-fPIC'"]}
x86_64-linux-gnu-gcc -pthread -DNDEBUG -g -fwrapv -O2 -Wall -Wstrict-prototypes -fno-strict-aliasing -Wdate-time -D_FORTIFY_SOURCE=2 -g -fstack-protector-strong -Wformat -Werror=format-security -fPIC -I/home/icarus/.local/lib/python2.7/site-packages/numpy/core/include -I/usr/include -I/usr/include/python2.7 -c nms/gpu_nms.cpp -o build/temp.linux-x86_64-2.7/nms/gpu_nms.o -Wno-unused-function
cc1plus: warning: command line option ‘-Wstrict-prototypes’ is valid for C/ObjC but not for C++
In file included from /home/icarus/.local/lib/python2.7/site-packages/numpy/core/include/numpy/ndarraytypes.h:1821:0,
                 from /home/icarus/.local/lib/python2.7/site-packages/numpy/core/include/numpy/ndarrayobject.h:18,
                 from /home/icarus/.local/lib/python2.7/site-packages/numpy/core/include/numpy/arrayobject.h:4,
                 from nms/gpu_nms.cpp:346:
/home/icarus/.local/lib/python2.7/site-packages/numpy/core/include/numpy/npy_1_7_deprecated_api.h:15:2: warning: #warning "Using deprecated NumPy API, disable it by " "#defining NPY_NO_DEPRECATED_API NPY_1_7_API_VERSION" [-Wcpp]
 #warning "Using deprecated NumPy API, disable it by " \
  ^
c++ -pthread -shared -Wl,-O1 -Wl,-Bsymbolic-functions -Wl,-Bsymbolic-functions -Wl,-z,relro -fno-strict-aliasing -DNDEBUG -g -fwrapv -O2 -Wall -Wstrict-prototypes -Wdate-time -D_FORTIFY_SOURCE=2 -g -fstack-protector-strong -Wformat -Werror=format-security -Wl,-Bsymbolic-functions -Wl,-z,relro -Wdate-time -D_FORTIFY_SOURCE=2 -g -fstack-protector-strong -Wformat -Werror=format-security build/temp.linux-x86_64-2.7/nms/nms_kernel.o build/temp.linux-x86_64-2.7/nms/gpu_nms.o -lcudart -o /home/icarus/oliver/lighttrack/lib/nms/gpu_nms.so
rm -rf build

求教pip安装torchlight失败

pip install torchlight==1.0出现
Could not find a version that satisfies the requirement torchlight==1.0 (from versions: 0.0.1)
No matching distribution found for torchlight==1.0
python demo_video_mobile.py出现
cannot import name 'str2bool'

Adding HrNet as pose estimator ?

Is it possible to add HrNet hrnet or posenet posenet into the tracking module ??

dockerfile

Thank for this project.

Can you provide a dockerfile to easily test lighttrack ?

AttributeError: 'Tensor' object has no attribute 'argsort'

how to slove the problem?

Evaluation on PoseTrack 2018 ?

Thanks for your great work.
I try to evaluation on PoseTrack 2018 with jsonformat_std_to_posetrack18.py -e 0.4 -d lighttrack -m track -f 17 -r 0.80, but, why the option "-f" is 17?

ModuleNotFoundError: No Module named nms.gpu_nms

now I have a problem is that :Traceback (most recent call last):
File "demo_video_mobile.py", line 28, in
from nms.gpu_nms import gpu_nms
ModuleNotFoundError: No module named 'nms.gpu_nms'

ModuleNotFoundError: No module named 'tfflat'

now I have a problem is that : "ModuleNotFoundError: No module named 'tfflat'",and I searched a lot and couldn't find a solution

(torchg) D:\desktop\myFramework\lighttrack>python demo_video_mobile.py
Traceback (most recent call last):
File "demo_video_mobile.py", line 17, in
from network_mobile_deconv import Network
File "D:\desktop\myFramework\lighttrack\network_mobile_deconv.py", line 14, in
from HPE.config import cfg
File "D:\desktop\myFramework\lighttrack\HPE\config.py", line 109, in
from tfflat.utils import add_pypath, make_link, make_dir
ModuleNotFoundError: No module named 'tfflat'

do I need install torch if I have installed torchlight?

I have installed torchlight and TF, when I run demo_video_mobile.py, I got "ModuleNotFoundError: No module named 'torch' ", Do I need install pytorch as well even I have installed torchlight?

TypeError: Error converting shape to a TensorShape: int() argument must be a string, a bytes-like object or a number, not 'tuple'

I run python demo_video_mobile.py ,but got error 'TypeError: Error converting shape to a TensorShape: int() argument must be a string, a bytes-like object or a number, not 'tuple'.'
tensorflow:1.12
python:3.6.8
pytorch:1.1.0

OpenSVAI can't be opened in China, how can I get it?

There are many files about openSVAI in the code.