CornerNet-Lite
CornerNet-Lite copied to clipboard
princeton-vl
Can it support in Windows10?
We only tested CornerNet-Lite on Linux and we don't have a Windows machine so I am not sure if the code will run properly on Windows 10.
D:\Python\Python36\lib\site-packages\torch\lib\include\torch\csrc\api\include\torch/torch.h(7): fatal error C1021: invalid preprocessor command 'warning' error: command 'C:\Program Files (x86)\Microsoft Visual Studio 14.0\VC\BIN\x86_amd64\cl.exe' failed with exit status 2
D:\Python\Python36\lib\site-packages\torch\lib\include\torch\csrc\api\include\torch/torch.h(7): fatal error C1021: invalid preprocessor command 'warning' error: command 'C:\Program Files (x86)\Microsoft Visual Studio 14.0\VC\BIN\x86_amd64\cl.exe' failed with exit status 2
I have met the same problem. And couldn't solve it. Anyone can help me? tks.
error: command 'C:\Program Files (x86)\Microsoft Visual Studio 14.0\VC\BIN\x86_amd64\cl.exe' failed with exit status 2
okay. I find the problem and solve it from the Pull requests. modify the four cpp files(bottom_pool.cpp, left_pool.cpp, top_pool.cpp, right_pool.cpp). Just change <torch/torch.h> to <torch/extension.h> will fix it.
okay. I find the problem and solve it from the Pull requests. modify the four cpp files(bottom_pool.cpp, left_pool.cpp, top_pool.cpp, right_pool.cpp). Just change <torch/torch.h> to <torch/extension.h> will fix it.
=========== 1.modify the four cpp files(bottom_pool.cpp, left_pool.cpp, top_pool.cpp, right_pool.cpp). Just change <torch/torch.h> to <torch/extension.h> 2.Also, insted install into user directory, we can build develop insted : #python setup.py install --user python setup.py build develop
THX!!!!
I have met the same problem. And when I modify the four cpp files(bottom_pool.cpp, left_pool.cpp, top_pool.cpp, right_pool.cpp). Just change <torch/torch.h> to <torch/extension.h>,the other error occured: command 'C:\Program Files (x86)\Microsoft Visual Studio 14.0\VC\BIN\x86_amd64\link.exe' failed with exit status 1171. Can you tell me how to solve it?
I have met the same problem. And when I modify the four cpp files(bottom_pool.cpp, left_pool.cpp, top_pool.cpp, right_pool.cpp). Just change <torch/torch.h> to <torch/extension.h>,the other error occured: command 'C:\Program Files (x86)\Microsoft Visual Studio 14.0\VC\BIN\x86_amd64\link.exe' failed with exit status 1171. Can you tell me how to solve it?
Compiling NMS Compile the NMS code which are originally from Faster R-CNN and Soft-NMS.
cd <CornerNet-Lite dir>/core/external make
D:\Tensorflow\pytorch\CornerNet-Lite\core\external>make python setup.py build_ext --inplace Compiling bbox.pyx because it changed. Compiling nms.pyx because it changed. [1/2] Cythonizing bbox.pyx D:\Program Files\Python368\lib\site-packages\Cython\Compiler\Main.py:367: FutureWarning: Cython directive 'language_level' not set, using 2 for now (Py2). This will change in a later release! File: D:\Tensorflow\pytorch\CornerNet-Lite\core\external\bbox.pyx tree = Parsing.p_module(s, pxd, full_module_name) [2/2] Cythonizing nms.pyx D:\Program Files\Python368\lib\site-packages\Cython\Compiler\Main.py:367: FutureWarning: Cython directive 'language_level' not set, using 2 for now (Py2). This will change in a later release! File: D:\Tensorflow\pytorch\CornerNet-Lite\core\external\nms.pyx tree = Parsing.p_module(s, pxd, full_module_name) running build_ext building 'bbox' extension creating build creating build\temp.win-amd64-3.6 creating build\temp.win-amd64-3.6\Release C:\Program Files (x86)\Microsoft Visual Studio\2017\Enterprise\VC\Tools\MSVC\14.16.27023\bin\HostX86\x64\cl.exe /c /nologo /Ox /W3 /GL /DNDEBUG /MD "-ID:\Program Files\Python368\lib\site-packages\numpy\core\include" "-ID:\Program Files\Python368\include" "-ID:\Program Files\Python368\include" "-IC:\Program Files (x86)\Microsoft Visual Studio\2017\Enterprise\VC\Tools\MSVC\14.16.27023\ATLMFC\include" "-IC:\Program Files (x86)\Microsoft Visual Studio\2017\Enterprise\VC\Tools\MSVC\14.16.27023\include" "-IC:\Program Files (x86)\Windows Kits\NETFXSDK\4.6.1\include\um" "-IC:\Program Files (x86)\Windows Kits\10\include\10.0.17763.0\ucrt" "-IC:\Program Files (x86)\Windows Kits\10\include\10.0.17763.0\shared" "-IC:\Program Files (x86)\Windows Kits\10\include\10.0.17763.0\um" "-IC:\Program Files (x86)\Windows Kits\10\include\10.0.17763.0\winrt" "-IC:\Program Files (x86)\Windows Kits\10\include\10.0.17763.0\cppwinrt" /Tcbbox.c /Fobuild\temp.win-amd64-3.6\Release\bbox.obj -Wno-cpp -Wno-unused-function cl: 命令行 error D8021 :无效的数值参数“/Wno-cpp” error: command 'C:\Program Files (x86)\Microsoft Visual Studio\2017\Enterprise\VC\Tools\MSVC\14.16.27023\bin\HostX86\x64\cl.exe' failed with exit status 2 make: *** [Makefile:2: all] Error 1
================= Anyone solved bbox and nms-gpu in win10？
https://blog.csdn.net/qq_41895190/article/details/82877883
cl: 命令行 error D8021 :无效的数值参数“/Wno-cpp”
============== okay. I find the problem and solve it.
是因为pycocotools不支持windows。 Because pycocotools does not support windows. pip install git+https://github.com/philferriere/cocoapi.git#subdirectory=PythonAPI 打印出如下信息，已经安装成功了。
C:\Users\TengLong>pip install git+https://github.com/philferriere/cocoapi.git#subdirectory=PythonAPI Looking in indexes: https://pypi.tuna.tsinghua.edu.cn/simple Collecting git+https://github.com/philferriere/cocoapi.git#subdirectory=PythonAPI Cloning https://github.com/philferriere/cocoapi.git to c:\users\tenglong\appdata\local\temp\pip-req-build-pkksskme Requirement already satisfied (use --upgrade to upgrade): pycocotools==2.0 from git+https://github.com/philferriere/cocoapi.git#subdirectory=PythonAPI in d:\program files\python368\lib\site-packages Building wheels for collected packages: pycocotools Building wheel for pycocotools (setup.py) ... done Stored in directory: C:\Users\TengLong\AppData\Local\Temp\pip-ephem-wheel-cache-__zczn1n\wheels\69\2b\12\2fa959e49f73d26cff202c2f4e5079096c9c57c8a8509fd75c Successfully built pycocotools
================ "D:\Program Files\Python368\python.exe" D:/Tensorflow/pytorch/CornerNet-Lite/demo.py total parameters: 116969339 loading from D:\Tensorflow\pytorch\CornerNet-Lite\core..\cache\nnet\CornerNet_Saccade\CornerNet_Saccade_500000.pkl
python setup.py build_ext --inplace
Compiling NMS Compile the NMS code which are originally from Faster R-CNN and Soft-NMS.
cd /core/external
make D:\Tensorflow\pytorch\CornerNet-Lite\core\external>make python setup.py build_ext --inplace Compiling bbox.pyx because it changed. Compiling nms.pyx because it changed. [1/2] Cythonizing bbox.pyx D:\Program Files\Python368\lib\site-packages\Cython\Compiler\Main.py:367: FutureWarning: Cython directive 'language_level' not set, using 2 for now (Py2). This will change in a later release! File: D:\Tensorflow\pytorch\CornerNet-Lite\core\external\bbox.pyx tree = Parsing.p_module(s, pxd, full_module_name) [2/2] Cythonizing nms.pyx D:\Program Files\Python368\lib\site-packages\Cython\Compiler\Main.py:367: FutureWarning: Cython directive 'language_level' not set, using 2 for now (Py2). This will change in a later release! File: D:\Tensorflow\pytorch\CornerNet-Lite\core\external\nms.pyx tree = Parsing.p_module(s, pxd, full_module_name) running build_ext building 'bbox' extension creating build creating build\temp.win-amd64-3.6 creating build\temp.win-amd64-3.6\Release C:\Program Files (x86)\Microsoft Visual Studio\2017\Enterprise\VC\Tools\MSVC\14.16.27023\bin\HostX86\x64\cl.exe /c /nologo /Ox /W3 /GL /DNDEBUG /MD "-ID:\Program Files\Python368\lib\site-packages\numpy\core\include" "-ID:\Program Files\Python368\include" "-ID:\Program Files\Python368\include" "-IC:\Program Files (x86)\Microsoft Visual Studio\2017\Enterprise\VC\Tools\MSVC\14.16.27023\ATLMFC\include" "-IC:\Program Files (x86)\Microsoft Visual Studio\2017\Enterprise\VC\Tools\MSVC\14.16.27023\include" "-IC:\Program Files (x86)\Windows Kits\NETFXSDK\4.6.1\include\um" "-IC:\Program Files (x86)\Windows Kits\10\include\10.0.17763.0\ucrt" "-IC:\Program Files (x86)\Windows Kits\10\include\10.0.17763.0\shared" "-IC:\Program Files (x86)\Windows Kits\10\include\10.0.17763.0\um" "-IC:\Program Files (x86)\Windows Kits\10\include\10.0.17763.0\winrt" "-IC:\Program Files (x86)\Windows Kits\10\include\10.0.17763.0\cppwinrt" /Tcbbox.c /Fobuild\temp.win-amd64-3.6\Release\bbox.obj -Wno-cpp -Wno-unused-function cl: 命令行 error D8021 :无效的数值参数“/Wno-cpp” error: command 'C:\Program Files (x86)\Microsoft Visual Studio\2017\Enterprise\VC\Tools\MSVC\14.16.27023\bin\HostX86\x64\cl.exe' failed with exit status 2 make: *** [Makefile:2: all] Error 1
================= Anyone solved bbox and nms-gpu in win10？
**
change all extra_compile_args=["-Wno-cpp", "-Wno-unused-function"] to extra_compile_args={'gcc': ['/Qstd=c99']}
and run command ‘python setup.py build_ext --inplace’
but pycocotools for windows is required https://github.com/binhbumpro/pycocotools-window
**
modify core\external\setup.py change all extra_compile_args=["-Wno-cpp", "-Wno-unused-function"] to extra_compile_args={'gcc': ['/Qstd=c99']}
and run command ‘python setup.py build_ext --inplace’
but pycocotools for windows is required https://github.com/binhbumpro/pycocotools-window
win10 omg
RuntimeError: Dimension out of range (expected to be in range of [-4, 3], but got 2559800508416) (maybe_wrap_dim at ..\aten\src\ATen/core/WrapDimMinimal.h:18) (no backtrace available)
Anyone solved bbox and nms-gpu in win10？
modify core\external\setup.py change all extra_compile_args=["-Wno-cpp", "-Wno-unused-function"] to extra_compile_args={'gcc': ['/Qstd=c99']} and run command ‘python setup.py build_ext --inplace’ but pycocotools for windows is required https://github.com/binhbumpro/pycocotools-window
It can probably solve bbox-nms problem. But when running it in win10,I find CornerNet-Squeeze is slower than yolov3.
Anyone solved bbox and nms-gpu in win10？
I solve this soft-nms problem， I wirte a new function(pure python) replace the orign one. 我写了一个一个python 程序替代了原来的soft_nms
this function is a little different from orign one, it return two objects:
def soft_nms(boxes: np.ndarray, sigma: float = 0.5, Nt: float = 0.3, threshold: float = 0.001, method: int = 0): """ :param boxes: 要求两个维度[n,8] n个boxes :param sigma: :param Nt: iou的阈值 :param threshold: :param method: :return: """ n = np.shape(boxes)[0] # 多少个box if n == 0: return None, np.zeros((0, 5))
for i in range(n):
# 遍历boxes数组中的每个box, 指针每指向一个box后,
# 这个box会和数组后面的全部box(index比他大的box)做比较, 找到得分最大的box
# 用最大box坐标信息和得分与当前指针位置的box的坐标信息和得分,互相交换
# 然后指针后移开始计算下一个box
maxscore = boxes[i, 4] # index = 4 的位置上是 box的得分
maxpos = i # 最大得分的位置
tx1 = boxes[i, 0]
ty1 = boxes[i, 1]
tx2 = boxes[i, 2]
ty2 = boxes[i, 3]
ts = boxes[i, 4]
pos = i + 1
# get max box, 通过while循环, 找到index:i与boxes数组后面部分的最大得分的box的index和最大得分
while pos < n:
if maxscore < boxes[pos, 4]:
maxscore = boxes[pos, 4]
maxpos = pos
pos = pos + 1
# add max box as a detection,
# 把最大得分的box信息和当前指针位置的box信息互相交换
boxes[i, 0] = boxes[maxpos, 0]
boxes[i, 1] = boxes[maxpos, 1]
boxes[i, 2] = boxes[maxpos, 2]
boxes[i, 3] = boxes[maxpos, 3]
boxes[i, 4] = boxes[maxpos, 4]
# swap ith box with position of max box
boxes[maxpos, 0] = tx1
boxes[maxpos, 1] = ty1
boxes[maxpos, 2] = tx2
boxes[maxpos, 3] = ty2
boxes[maxpos, 4] = ts
pos = i + 1 # 重置pos 为i的下一个index
while pos < n:
x1 = boxes[pos, 0]
y1 = boxes[pos, 1]
x2 = boxes[pos, 2]
y2 = boxes[pos, 3]
s = boxes[pos, 4]
area = (x2 - x1 + 1) * (y2 - y1 + 1) #
iw = (min(tx2, x2) - max(tx1, x1) + 1) #
if iw > 0: # iw >0,ih>0 pos位置的box与i位置的box是相交的
ih = (min(ty2, y2) - max(ty1, y1) + 1) #
if ih > 0:
ua = float((tx2 - tx1 + 1) * (ty2 - ty1 + 1) + area - iw * ih)
#
ov = iw * ih / ua # iou between max box and detection box iou
if method == 1: # linear
if ov > Nt:
weight = 1 - ov
else:
weight = 1
elif method == 2: # gaussian 默认选择
weight = np.exp(-(ov * ov) / sigma)
else: # original NMS
if ov > Nt:
weight = 0
else:
weight = 1
# weight
# 对pos 位置的box的得分进行修改
boxes[pos, 4] = weight * boxes[pos, 4]
# if box score falls below threshold, discard the box by swapping with last box
# update n 如果box的得分经过抑制后低于设置的阈值,抛弃这个box, 使用数组中的最后一个box的信息做替换
if boxes[pos, 4] < threshold:
boxes[pos, 0] = boxes[n - 1, 0]
boxes[pos, 1] = boxes[n - 1, 1]
boxes[pos, 2] = boxes[n - 1, 2]
boxes[pos, 3] = boxes[n - 1, 3]
boxes[pos, 4] = boxes[n - 1, 4]
n = n - 1
pos = pos - 1
pos = pos + 1
keep = [i for i in range(n)]
return keep, boxes
and you need change code(cornernet_saccade.py) in the same place where function soft_nms is used. .......... top_bboxes = {} # {类别id:boxes(8个坐标的)} for j in range(categories): keep_inds = (classes == j) top_bboxes[j + 1] = detections[keep_inds][:, 0:7].astype(np.float32) # 预测的最后调用了 soft_nms # 这里需要修改, 传入的是8个坐标的boxes keep_inds, after_nms_boxes = soft_nms(top_bboxes[j + 1], Nt=nms_threshold, method=nms_algorithm, sigma=0.7) # print(keep_inds) # print(after_nms_boxes) # print(type(keep_inds)) # print("===============")
if keep_inds is not None:
one_boxes = []
for one_keep_index in keep_inds:
one_box = after_nms_boxes[one_keep_index]
one_boxes.append(one_box[0:5])
else:
one_boxes = after_nms_boxes
# top_bboxes[j + 1] = top_bboxes[j + 1][keep_inds, 0:5] # 4个坐标 + 上点和下点的平均分类得分
top_bboxes[j + 1] = np.array(one_boxes) # 4个坐标 + 上点和下点的平均分类得分
.......
We only tested CornerNet-Lite on Linux and we don't have a Windows machine so I am not sure if the code will run properly on Windows 10. Hello,when i python train.py,have a error. cfg_file = os.path.join(system_configs.config_dir, args.cfg_file + ".json") TypeError: unsupported operand type(s) for +: 'NoneType' and 'str' Can you help me?thank you.
