Pytracking代码运行笔记

曾经终败给现在 2022-01-20 22:31 240阅读 0赞

https://github.com/visionml/pytracking
    
    1.git clone https://github.com/visionml/pytracking.git
    2.cd pytracking/
    3.git submodule update --init  
    4.
    bash install.sh conda_install_path pytracking
    conda_install_path替换为：
    which conda
    输出anaconda3的路径
    即为：
    bash install.sh /data/anaconda3 pytracking
    特别的，关注一下install.sh强大的环境安装脚本文件
    ！！！应用脚本安装不成功（conda环境丢失），于是手动创建环境，手动安装
    
    conda install -y XXX 表示确认安装，确认创建
    当执行：python -c "XXX"时，XXX表示要执行的命令，例如：
    输入：python -c "import os; print(os.sys.executable)" 
    结果为：/usr/bin/python
    
    
    5.
    需要翻墙下载的两个权重模型：
    https://docs.google.com/uc?export=download&id=1aWC4waLv_te-BULoy0k-n_zS-ONms21S    可能不行
    https://docs.google.com/uc?export=download&id=1ZTdQbZ1tyN27UIwUnUrjHChQb5ug2sxr   真垃圾不行！
    https://drive.google.com/drive/folders/1WVhJqvdu-_JG1U-V0IqfxTUa1SBPnL0O   可以了！
    
    
    6.
    运行：python run_webcam.py atom default时
    出现错误：EOFError: Ran out of input 
    经过检索发现经常会在load文件不成功时出现该错误，于是去查看了模型文件，发现模型文件为空（0byte），则确定是模型文件出现问题，则翻墙自行下载模型文件，下载文件出错。
    出现错误：
    No matching checkpoint file found
    需要翻墙下载正确的权重模型。
    解决方案：
    多翻一翻issues中的问题和解答，总是有帮助的。
    
    不可以直接对下载的文件直接进行解压，会报错，因为他是一个torch的模型，需要用torch.load()来导入该文件。
    记得把文件名atom_default.pth.tar改名为：atom_default.pth（直接重命名即可）
    
    7.
    python run_tracker.py tracker_name parameter_name --dataset_name dataset_name --sequence sequence --debug debug --threads threads
    改为：
    python run_tracker.py atom default --dataset_name otb --sequence Soccer --debug debug --threads threads
    改为：
    python run_tracker.py atom default --dataset otb --sequence Soccer
    
    
    python run_experiment.py experiment_module experiment_name --dataset_name dataset_name --sequence sequence --debug debug --threads threads
    改为：
    python run_experiment.py  myexperiments  atom_nfs_uav --dataset otb --sequence Soccer
    改为：
    python run_experiment.py  myexperiments  atom_nfs_uav
    
    8.
    在成功解决No matching checkpoint file found问题之后，运行7中的两条命令都报错：
    RuntimeError: Attempting to deserialize object on a CUDA device but torch.cuda.is_available() is False. If you are running on a CPU-only machine, please use torch.load with map_location='cpu' to map your storages to the CPU.
    解决方法：
    出错语句：、
    torch.load(model_file)
    改为：
    model = torch.load(model_path, map_location='cpu')
    
    9.
    解决完问题8之后，出现问题：pytorch框架下报错cuda驱动版本不满足cuda运行版本。
    问题日志如下：
    log:
    (pytracking) 09534@09534:/data/pzn/pzncode/pytracking/pytracking$ python run_webcam.py atom default
    VIDEOIO ERROR: V4L: can't open camera by index 0
    No CUDA runtime is found, using CUDA_HOME='/usr/local/cuda'
    Using /tmp/torch_extensions as PyTorch extensions root...
    Detected CUDA files, patching ldflags
    Emitting ninja build file /tmp/torch_extensions/_prroi_pooling/build.ninja...
    Building extension module _prroi_pooling...
    ninja: no work to do.
    Loading extension module _prroi_pooling...
    Traceback (most recent call last):
      File "run_webcam.py", line 35, in <module>
        main()
      File "run_webcam.py", line 31, in main
        run_webcam(args.tracker_name, args.tracker_param, args.debug)
      File "run_webcam.py", line 20, in run_webcam
        tracker.run_webcam(debug)
      File "../pytracking/evaluation/tracker.py", line 94, in run_webcam
        tracker.track_webcam()
      File "../pytracking/tracker/base/basetracker.py", line 179, in track_webcam
        self.initialize_features()
      File "../pytracking/tracker/atom/atom.py", line 19, in initialize_features
        self.params.features.initialize()
      File "../pytracking/features/extractor.py", line 16, in initialize
        f.initialize()
      File "../pytracking/features/deep.py", line 98, in initialize
        self.net.cuda()
      File "/data/anaconda3/envs/pytracking/lib/python3.7/site-packages/torch/nn/modules/module.py", line 265, in cuda
        return self._apply(lambda t: t.cuda(device))
      File "/data/anaconda3/envs/pytracking/lib/python3.7/site-packages/torch/nn/modules/module.py", line 193, in _apply
        module._apply(fn)
      File "/data/anaconda3/envs/pytracking/lib/python3.7/site-packages/torch/nn/modules/module.py", line 193, in _apply
        module._apply(fn)
      File "/data/anaconda3/envs/pytracking/lib/python3.7/site-packages/torch/nn/modules/module.py", line 199, in _apply
        param.data = fn(param.data)
      File "/data/anaconda3/envs/pytracking/lib/python3.7/site-packages/torch/nn/modules/module.py", line 265, in <lambda>
        return self._apply(lambda t: t.cuda(device))
      File "/data/anaconda3/envs/pytracking/lib/python3.7/site-packages/torch/cuda/__init__.py", line 162, in _lazy_init
        _check_driver()
      File "/data/anaconda3/envs/pytracking/lib/python3.7/site-packages/torch/cuda/__init__.py", line 91, in _check_driver
        of the CUDA driver.""".format(str(torch._C._cuda_getDriverVersion())))
    AssertionError:
    The NVIDIA driver on your system is too old (found version 9000).
    Please update your GPU driver by downloading and installing a new
    version from the URL: http://www.nvidia.com/Download/index.aspx
    Alternatively, go to: https://pytorch.org to install
    a PyTorch version that has been compiled with your version
    of the CUDA driver.
    
    问题原因分析：
    每个pytorch版本都有对应的cuda版本，在安装pytorch时，如果不指定版本，则会首选安装最近版本的的pytorch，这样对应的cuda版本会很高，和本机的cuda版本以及NVIDIA驱动版本不相符，则会报错英伟达驱动太老了，建议更新英伟达驱动。
    解决方案：其实不更新英伟达驱动也是可以的，解决方案就是降低pytorch的版本，即卸载原始的pytorch，安装指定低版本的pytorch即可解决该问题。
    （1）通过该命令可以查看pytorch的当前版本
    import torch
    torch.__version__
    torch.cuda.is_available()
    （2）查看本机的cuda版本
    #查看Linux server安装的cuda版本
    #切换到/usr/local/cuda/samples/1_Utilities/deviceQuery然后运行
    make
    ./deviceQuery
    （3）对应本机的cuda版本选择合适版本的pytorch
    conda/pip3 uninstall pytorch
    conda/pip3 install [pytorch-version-link]
    
    09534服务器的cuda版本为9.0，其对应的pytorch版本为：0.4.0
    即运行以下命令即可:
    conda uninstall pytorch
    conda install pytorch=0.4.0
    
    10.
    出现问题：缺少高版本pytorch中的扩展包的头文件，所以还是不能选用如此低版本的pytorch
    日志文件如下：
    logs:
    
    (pytracking) 09534@09534:/data/pzn/pzncode/pytracking/pytracking$ python run_webcam.py atom default
    VIDEOIO ERROR: V4L: can't open camera by index 0
    Using /tmp/torch_extensions as PyTorch extensions root...
    Detected CUDA files, patching ldflags
    Emitting ninja build file /tmp/torch_extensions/_prroi_pooling/build.ninja...
    Building extension module _prroi_pooling...
    Traceback (most recent call last):
      File "/data/anaconda3/envs/pytracking/lib/python3.6/site-packages/torch/utils/cpp_extension.py", line 576, in _build_extension_module
        ['ninja', '-v'], stderr=subprocess.STDOUT, cwd=build_directory)
      File "/data/anaconda3/envs/pytracking/lib/python3.6/subprocess.py", line 356, in check_output
        **kwargs).stdout
      File "/data/anaconda3/envs/pytracking/lib/python3.6/subprocess.py", line 438, in run
        output=stdout, stderr=stderr)
    subprocess.CalledProcessError: Command '['ninja', '-v']' returned non-zero exit status 1.
    
    During handling of the above exception, another exception occurred:
    
    Traceback (most recent call last):
      File "run_webcam.py", line 35, in <module>
        main()
      File "run_webcam.py", line 31, in main
        run_webcam(args.tracker_name, args.tracker_param, args.debug)
      File "run_webcam.py", line 20, in run_webcam
        tracker.run_webcam(debug)
      File "../pytracking/evaluation/tracker.py", line 94, in run_webcam
        tracker.track_webcam()
      File "../pytracking/tracker/base/basetracker.py", line 179, in track_webcam
        self.initialize_features()
      File "../pytracking/tracker/atom/atom.py", line 19, in initialize_features
        self.params.features.initialize()
      File "../pytracking/features/extractor.py", line 16, in initialize
        f.initialize()
      File "../pytracking/features/deep.py", line 95, in initialize
        self.net, _ = load_network(net_path_full, backbone_pretrained=False)
      File "../ltr/admin/loading.py", line 70, in load_network
        net = net_constr.get()
      File "../ltr/admin/model_constructor.py", line 43, in get
        net_module = importlib.import_module(self.fun_module)
      File "/data/anaconda3/envs/pytracking/lib/python3.6/importlib/__init__.py", line 126, in import_module
        return _bootstrap._gcd_import(name[level:], package, level)
      File "<frozen importlib._bootstrap>", line 994, in _gcd_import
      File "<frozen importlib._bootstrap>", line 971, in _find_and_load
      File "<frozen importlib._bootstrap>", line 941, in _find_and_load_unlocked
      File "<frozen importlib._bootstrap>", line 219, in _call_with_frames_removed
      File "<frozen importlib._bootstrap>", line 994, in _gcd_import
      File "<frozen importlib._bootstrap>", line 971, in _find_and_load
      File "<frozen importlib._bootstrap>", line 955, in _find_and_load_unlocked
      File "<frozen importlib._bootstrap>", line 665, in _load_unlocked
      File "<frozen importlib._bootstrap_external>", line 678, in exec_module
      File "<frozen importlib._bootstrap>", line 219, in _call_with_frames_removed
      File "../ltr/models/bbreg/__init__.py", line 1, in <module>
        from .atom_iou_net import AtomIoUNet
      File "../ltr/models/bbreg/atom_iou_net.py", line 4, in <module>
        from ltr.external.PreciseRoIPooling.pytorch.prroi_pool import PrRoIPool2D
      File "../ltr/external/PreciseRoIPooling/pytorch/prroi_pool/__init__.py", line 12, in <module>
        from .prroi_pool import *
      File "../ltr/external/PreciseRoIPooling/pytorch/prroi_pool/prroi_pool.py", line 14, in <module>
        from .functional import prroi_pool2d
      File "../ltr/external/PreciseRoIPooling/pytorch/prroi_pool/functional.py", line 22, in <module>
        verbose=True
      File "/data/anaconda3/envs/pytracking/lib/python3.6/site-packages/torch/utils/cpp_extension.py", line 501, in load
        _build_extension_module(name, build_directory)
      File "/data/anaconda3/envs/pytracking/lib/python3.6/site-packages/torch/utils/cpp_extension.py", line 582, in _build_extension_module
        name, error.output.decode()))
    RuntimeError: Error building extension '_prroi_pooling': [1/2] c++ -MMD -MF prroi_pooling_gpu.o.d -DTORCH_EXTENSION_NAME=_prroi_pooling -I/data/anaconda3/envs/pytracking/lib/python3.6/site-packages/torch/lib/include -I/data/anaconda3/envs/pytracking/lib/python3.6/site-packages/torch/lib/include/TH -I/data/anaconda3/envs/pytracking/lib/python3.6/site-packages/torch/lib/include/THC -I/usr/local/cuda/include -I/data/anaconda3/envs/pytracking/include/python3.6m -fPIC -std=c++11 -c /data/pzn/pzncode/pytracking/ltr/external/PreciseRoIPooling/pytorch/prroi_pool/src/prroi_pooling_gpu.c -o prroi_pooling_gpu.o
    FAILED: prroi_pooling_gpu.o
    c++ -MMD -MF prroi_pooling_gpu.o.d -DTORCH_EXTENSION_NAME=_prroi_pooling -I/data/anaconda3/envs/pytracking/lib/python3.6/site-packages/torch/lib/include -I/data/anaconda3/envs/pytracking/lib/python3.6/site-packages/torch/lib/include/TH -I/data/anaconda3/envs/pytracking/lib/python3.6/site-packages/torch/lib/include/THC -I/usr/local/cuda/include -I/data/anaconda3/envs/pytracking/include/python3.6m -fPIC -std=c++11 -c /data/pzn/pzncode/pytracking/ltr/external/PreciseRoIPooling/pytorch/prroi_pool/src/prroi_pooling_gpu.c -o prroi_pooling_gpu.o
    /data/pzn/pzncode/pytracking/ltr/external/PreciseRoIPooling/pytorch/prroi_pool/src/prroi_pooling_gpu.c:12:29: fatal error: torch/extension.h: No such file or directory
    compilation terminated.
    ninja: build stopped: subcommand failed.
    
    参考网站，使用该命令进行重新安装pytorch：
    conda install pytorch torchvision cudatoolkit=9.0 -c pytorch
    
    pytorch包下载链接：
    https://conda.anaconda.org/pytorch/linux-64/pytorch-1.1.0-py3.6_cuda9.0.176_cudnn7.5.1_0.tar.bz2
    
    查看conda的源：
    conda config --show channels
    删除conda的指定源：
    conda config --remove channels 'https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free/'
    conda换回默认源：
    conda config --remove-key channels
    
    由于使用conda直接安装，又没有添加合适的源会使得下载速度过慢，或者是下载不下来的情况，尝试把下载好的conda安装的pytorch包放到路径anaconda3/pkgs下并解压，再用conda继续安装，但是没有成功；
    所以采用pip直接对pytorch进行安装，从官网上可知： https://pytorch.org/
    
    当cuda的版本为9.0时，直接使用命令：pip install torch torchvision即可。
    安装好的pytorch版本仍为：1.1.0
    即满足高版本的pytorch和低版本的cuda driver的需求。
    
    
    11.
    （1）
    python run_webcam.py atom default因为调用不了摄像头，不能运行成功
    （2）
    python run_tracker.py atom default --dataset otb --sequence Soccer  在序列上运行成功
    结果保存在：pytracking/pytracking/tracking_results/atom/default/下
    （3）
    python run_experiment.py  myexperiments  atom_nfs_uav 没有下载这个数据集
    python run_experiment.py  myexperiments  otb_test  在otb数据集上运行成功
    
    代码修改：
    （1）修改了pytracking/pytracking/evaluation/local.py文件中的第12行：otb_path的路径：
    12 line:  (fix)
        #settings.otb_path = ''#pang_comment
        settings.otb_path = '/data/pzn/pzncode/pytracking/pytracking/dataset/otb' #pang_add
    
    （2）为解决问题8，修改文件：
    pytracking/ltr/admin/loading.py文件中的第97行：
    97 line:  (fix)
        # Load network
        #checkpoint_dict = torch.load(path) #pang_comment
        checkpoint_dict = torch.load(path, map_location='cpu')  #pang_add
    
    （3）为了在整个数据集上的做测试，修改文件：
    pytracking/pytracking/experiments/myexperiments.py文件中的第20行：增加
    line 20:  (add)
    def otb_test():
        # Run ATOM and ECO on the UAV dataset
        trackers = [Tracker('atom', 'default', i) for i in range(1)]
    
        dataset = OTBDataset()
        return trackers, dataset
    
    为了测试在otb数据集上的结果和表现，结果保存在pytracking/pytracking/tracking_results/atom/default_000/下
    
    表现如下：
    performance:
    (pytracking) 09534@09534:/data/pzn/pzncode/pytracking/pytracking$ python run_experiment.py  myexperiments  otb_test
    Running:  myexperiments  otb_test
    Tracker: atom default 0 ,  Sequence: Basketball
    Using /tmp/torch_extensions as PyTorch extensions root...
    Detected CUDA files, patching ldflags
    Emitting ninja build file /tmp/torch_extensions/_prroi_pooling/build.ninja...
    Building extension module _prroi_pooling...
    ninja: no work to do.
    Loading extension module _prroi_pooling...
    /data/anaconda3/envs/pytracking/lib/python3.6/site-packages/torch/nn/functional.py:2539: UserWarning: Default upsampling behavior when mode=bilinear is changed to align_corners=False since 0.4.0. Please specify align_corners=True if the old behavior is desired. See the documentation of nn.Upsample for details.
      "See the documentation of nn.Upsample for details.".format(mode))
    FPS: 32.54937622204515
    Tracker: atom default 0 ,  Sequence: Biker
    FPS: 29.495022863933915
    Tracker: atom default 0 ,  Sequence: Bird1
    FPS: 25.06367000979083
    Tracker: atom default 0 ,  Sequence: Bird2
    FPS: 26.741711434101838
    Tracker: atom default 0 ,  Sequence: BlurBody
    FPS: 31.083897248157196
    Tracker: atom default 0 ,  Sequence: BlurCar1
    FPS: 26.793512016749418
    Tracker: atom default 0 ,  Sequence: BlurCar2
    FPS: 31.48350616426953
    Tracker: atom default 0 ,  Sequence: BlurCar3
    FPS: 23.017989835856113
    Tracker: atom default 0 ,  Sequence: BlurCar4
    FPS: 31.663243147349203
    Tracker: atom default 0 ,  Sequence: BlurFace
    FPS: 31.080425117194615
    Tracker: atom default 0 ,  Sequence: BlurOwl
    FPS: 39.805202856138095
    Tracker: atom default 0 ,  Sequence: Board
    FPS: 32.15119785229665
    Tracker: atom default 0 ,  Sequence: Bolt
    FPS: 30.41741483062557
    Tracker: atom default 0 ,  Sequence: Bolt2
    FPS: 27.695008347166777
    Tracker: atom default 0 ,  Sequence: Box
    FPS: 31.638685554108175
    Tracker: atom default 0 ,  Sequence: Boy
    FPS: 42.15040810493713
    Tracker: atom default 0 ,  Sequence: Car1
    FPS: 31.943427969784743
    Tracker: atom default 0 ,  Sequence: Car2
    FPS: 31.742681824728784
    Tracker: atom default 0 ,  Sequence: Car24
    FPS: 33.67028693347804
    Tracker: atom default 0 ,  Sequence: Car4
    FPS: 31.03038872679038
    Tracker: atom default 0 ,  Sequence: CarDark
    FPS: 31.433773656093244
    Tracker: atom default 0 ,  Sequence: CarScale
    FPS: 30.703222560704873
    Tracker: atom default 0 ,  Sequence: ClifBar
    FPS: 31.248336541226823
    Tracker: atom default 0 ,  Sequence: Coke
    FPS: 21.470539231995012
    Tracker: atom default 0 ,  Sequence: Couple
    FPS: 29.169828414895484
    Tracker: atom default 0 ,  Sequence: Coupon
    FPS: 30.588841735753782
    Tracker: atom default 0 ,  Sequence: Crossing
    FPS: 29.55670446900004
    Tracker: atom default 0 ,  Sequence: Crowds
    FPS: 30.849982099498806
    Tracker: atom default 0 ,  Sequence: Dancer
    FPS: 69.8099648101612
    Tracker: atom default 0 ,  Sequence: Dancer2
    FPS: 28.30845038463361
    Tracker: atom default 0 ,  Sequence: David
    FPS: 32.12054991984322
    Tracker: atom default 0 ,  Sequence: David2
    FPS: 26.53936763770796
    Tracker: atom default 0 ,  Sequence: David3
    FPS: 30.862871204517166
    Tracker: atom default 0 ,  Sequence: Deer
    FPS: 24.345640626854767
    Tracker: atom default 0 ,  Sequence: Diving
    FPS: 30.55935068992381
    Tracker: atom default 0 ,  Sequence: Dog
    FPS: 29.00565469024118
    Tracker: atom default 0 ,  Sequence: Dog1
    FPS: 32.18601409498473
    Tracker: atom default 0 ,  Sequence: Doll
    FPS: 34.56547014491602
    Tracker: atom default 0 ,  Sequence: DragonBaby
    FPS: 27.12983381961647
    Tracker: atom default 0 ,  Sequence: Dudek
    FPS: 35.61511205677792
    Tracker: atom default 0 ,  Sequence: FaceOcc1
    FPS: 37.368102206834166
    Tracker: atom default 0 ,  Sequence: FaceOcc2
    FPS: 30.597189242272954
    Tracker: atom default 0 ,  Sequence: Fish
    FPS: 30.500117899319648
    Tracker: atom default 0 ,  Sequence: FleetFace
    FPS: 32.01345618145981
    Tracker: atom default 0 ,  Sequence: Football
    FPS: 30.65345475240097
    Tracker: atom default 0 ,  Sequence: Football1
    FPS: 34.53106381179711
    Tracker: atom default 0 ,  Sequence: Freeman1
    FPS: 31.681660990731444
    Tracker: atom default 0 ,  Sequence: Freeman3
    FPS: 31.239276707496735
    Tracker: atom default 0 ,  Sequence: Freeman4
    FPS: 31.83163635300702
    Tracker: atom default 0 ,  Sequence: Girl
    FPS: 32.40600417328795
    Tracker: atom default 0 ,  Sequence: Girl2
    FPS: 34.62853357512583
    Tracker: atom default 0 ,  Sequence: Gym
    FPS: 31.82578986650551
    Tracker: atom default 0 ,  Sequence: Human2
    FPS: 32.35969507473828
    Tracker: atom default 0 ,  Sequence: Human3
    FPS: 33.00868927050219
    Tracker: atom default 0 ,  Sequence: Human4_2
    FPS: 31.779914383961763
    Tracker: atom default 0 ,  Sequence: Human5
    FPS: 26.76457291750291
    Tracker: atom default 0 ,  Sequence: Human6
    FPS: 30.07621386711268
    Tracker: atom default 0 ,  Sequence: Human7
    FPS: 29.776363205659308
    Tracker: atom default 0 ,  Sequence: Human8
    FPS: 30.458452170660994
    Tracker: atom default 0 ,  Sequence: Human9
    FPS: 31.171137137117235
    Tracker: atom default 0 ,  Sequence: Ironman
    FPS: 28.894834309056005
    Tracker: atom default 0 ,  Sequence: Jogging_1
    FPS: 31.315911864349236
    Tracker: atom default 0 ,  Sequence: Jogging_2
    FPS: 23.0590683486986
    Tracker: atom default 0 ,  Sequence: Jump
    FPS: 28.996989431815962
    Tracker: atom default 0 ,  Sequence: Jumping
    FPS: 31.921820766910752
    Tracker: atom default 0 ,  Sequence: KiteSurf
    FPS: 27.284018799499115
    Tracker: atom default 0 ,  Sequence: Lemming
    FPS: 34.35385842248308
    Tracker: atom default 0 ,  Sequence: Liquor
    FPS: 33.10012804656343
    Tracker: atom default 0 ,  Sequence: Man
    FPS: 28.914124182059002
    Tracker: atom default 0 ,  Sequence: Matrix
    FPS: 28.829349972770817
    Tracker: atom default 0 ,  Sequence: Mhyang
    FPS: 24.960922069047538
    Tracker: atom default 0 ,  Sequence: MotorRolling
    FPS: 28.135487666727094
    Tracker: atom default 0 ,  Sequence: MountainBike
    FPS: 29.160919907903704
    Tracker: atom default 0 ,  Sequence: Panda
    FPS: 31.392217666871804
    Tracker: atom default 0 ,  Sequence: RedTeam
    FPS: 30.40206547995665
    Tracker: atom default 0 ,  Sequence: Rubik
    FPS: 29.325685533024092
    Tracker: atom default 0 ,  Sequence: Shaking
    FPS: 29.460189493257424
    Tracker: atom default 0 ,  Sequence: Singer1
    FPS: 30.165988137162195
    Tracker: atom default 0 ,  Sequence: Singer2
    FPS: 31.132716959934108
    Tracker: atom default 0 ,  Sequence: Skater
    FPS: 29.393834645033326
    Tracker: atom default 0 ,  Sequence: Skater2
    FPS: 23.177624658689975
    Tracker: atom default 0 ,  Sequence: Skating1
    FPS: 32.60175017668689
    Tracker: atom default 0 ,  Sequence: Skating2_1
    FPS: 30.1289522986984
    Tracker: atom default 0 ,  Sequence: Skating2_2
    FPS: 33.39295187412872
    Tracker: atom default 0 ,  Sequence: Skiing
    FPS: 27.071525005404524
    Tracker: atom default 0 ,  Sequence: Soccer
    FPS: 33.245764078114185
    Tracker: atom default 0 ,  Sequence: Subway
    FPS: 29.68875880900868
    Tracker: atom default 0 ,  Sequence: Surfer
    FPS: 30.516335378652023
    Tracker: atom default 0 ,  Sequence: Suv
    FPS: 31.888240715424317
    Tracker: atom default 0 ,  Sequence: Sylvester
    FPS: 31.71966007563366
    Tracker: atom default 0 ,  Sequence: Tiger1
    FPS: 30.578924915050074
    Tracker: atom default 0 ,  Sequence: Tiger2
    FPS: 22.495691874633252
    Tracker: atom default 0 ,  Sequence: Toy
    FPS: 30.445012542208687
    Tracker: atom default 0 ,  Sequence: Trans
    FPS: 28.162139867736514
    Tracker: atom default 0 ,  Sequence: Trellis
    FPS: 39.60974413799816
    Tracker: atom default 0 ,  Sequence: Twinnings
    FPS: 31.400159152219338
    Tracker: atom default 0 ,  Sequence: Vase
    FPS: 29.81456277601726
    Tracker: atom default 0 ,  Sequence: Walking
    FPS: 24.093669172186974
    Tracker: atom default 0 ,  Sequence: Walking2
    FPS: 31.54828275676745
    Tracker: atom default 0 ,  Sequence: Woman
    FPS: 32.121155275556404
    Done
    
    （4）为了得到运算的精度，则需要把生成的box文件保存成struct s{bb}的mat形式，特此修改代码：
    /data/pzn/pzncode/pytracking/pytracking/evaluation/running.py
    
    line 6:   (add)
    import scipy.io as scio #pang_add
    
    line 32:    (add)
    tracked_bb_yuan = tracked_bb  #pang_add
    tracked_bb = np.array(tracked_bb).astype(int)
    exec_times = np.array(exec_times).astype(float)
    
    line 45:   (add)
            np.savetxt(results_path, tracked_bb, delimiter='\t', fmt='%d')
            np.savetxt(times_path, exec_times, delimiter='\t', fmt='%f')
            #pang_add
            s = {'bb': tracked_bb_yuan}  #pang_add
            mat_path_pang = '{}_otb50.mat'.format(base_results_path)  #pang_add
            print('mat_path_pang=============')  #pang_add
            print(mat_path_pang)  #pang_add
            scio.savemat(mat_path_pang, {'s':s}, appendmat = True, do_compression = True)  #pang_add
    
    运行：python run_experiment.py  myexperiments  otb_test
    结果保存在pytracking/pytracking/tracking_results/atom/default_000/下。
    
    
    
    在uav123数据集上进行测试，以获得测试的结果：
    python run_experiment.py  myexperiments  uav_test
    
    在drone2018数据集上进行测试，以获得测试的结果：
    python run_experiment.py  myexperiments  uav_test
    
    由于Drone2018数据集的格式和uav数据集格式比较类似，则采用变动数据集格式的方式，需要在uav123数据集的基础上增加如下改变：
    (0)test.py dealwith the dataset change to list
    
    (1)/data/pzn/pzncode/pytracking/pytracking/evaluation/local.py
    line 17: (fix)
    
        #settings.uav_path = '' #pang_comment
        #settings.uav_path = '/data/pzn/dataset_pang/UAV123' #pang_add #uav123
        settings.uav_path = '/data/pzn/dataset_pang/visdrone2018' #pang_add #drone2018
    
    (2)/data/pzn/pzncode/pytracking/pytracking/evaluation/running.py
    line 42: (fix)
            #mat_path_pang = '{}_otb50.mat'.format(base_results_path)  #pang_add
            #mat_path_pang = '{}_uav123.mat'.format(base_results_path)  #pang_add #uav123
            mat_path_pang = '{}_drone2018.mat'.format(base_results_path)  #pang_add #drone2018
    (3)
    操作过程中遇到错误：（错误日志如下：）
    error:(log)
    OpenCV(4.1.0) /io/opencv/modules/imgproc/src/color.cpp:182: error: (-215:Assertion failed) !_src.empty() in function 'cvtColor'
    
    出现错误的原因是：读取不到图片数据，才会报以上错误；
    解决方案：
    操作：先打印原始读入数据的路径，查看图片路径是否正确，发现果然不正确，在图片的文件格式是img%07d.jpg，原始获取图片文件的格式为%010d.jpg，所以要对格式进行重新调整。
    
    /data/pzn/pzncode/pytracking/pytracking/evaluation/uavdataset.py
    line 39:
    
            #frames = ['{base_path}/{sequence_path}/{frame:0{nz}}.{ext}'.format(base_path=self.base_path,
            #    sequence_path=sequence_path, frame=frame_num, nz=nz, ext=ext) for frame_num in range(start_frame+init_omit, end_frame+1)] #pang_comment
    
            frames = ['{base_path}/{sequence_path}/{img}{frame:0{nz}}.{ext}'.format(base_path=self.base_path,
                sequence_path=sequence_path, img='img', frame=frame_num, nz=nz, ext=ext) for frame_num in range(start_frame+init_omit, end_frame+1)] #pang_fix
    
    line 64:
    sequence_info_list = [] fill it from step (0).

建议同GitHub上pytracking代码的readme共同参考。