FENet: Focusing Enhanced Network for Lane Detection 调试报错记录
github项目链接
问题1：
指令：
(fenetv1) ws@sun:~$ python /home/ws/CoodWorkRun/FENet-v1/setup.py build develop
报错
RuntimeError:
The detected CUDA version (9.1) mismatches the version that was used to compile
PyTorch (11.3). Please make sure to use the same CUDA versions.
解决办法：
# 命令行终端先修正CUDA环境变量
export CUDA_PATH=/usr/local/cuda-11.3
export LD_LIBRARY_PATH=${LD_LIBRARY_PATH}:/usr/local/cuda-11.3/lib64
export PATH=${PATH}:/usr/local/cuda-11.3/bin
# 执行
python /home/ws/CoodWorkRun/FENet-v1/setup.py build develop
问题2：
指令：
(fenetv1) ws@sun:~$ python /home/ws/CoodWorkRun/FENet-v1/setup.py build develop
报错
# 1
ImportError: cannot import name 'nms_impl' from partially initialized module 'fenet.ops' (most likely due to a circular import) (/home/ws/CoodWorkRun/FENet-v1/fenet/ops/__init__.py)
# 2
copying build/lib.linux-x86_64-cpython-38/fenet/ops/nms_impl.cpython-38-x86_64-linux-gnu.so -> fenet/ops
error: could not create 'fenet/ops/nms_impl.cpython-38-x86_64-linux-gnu.so': No such file or directory解决办法：
尝试手动将.so文件复制到目标目录
build/lib.linux-x86_64-cpython-38/fenet/ops/nms_impl.cpython-38-x86_64-linux-gnu.so
问题3：
指令：
(fenetv2) ws@sun:~$ python /home/ws/CoodWorkRun/FENet-v1/main.py
报错
Traceback (most recent call last):File "/home/ws/CoodWorkRun/FENet-v1/main.py", line 62, in <module>main()File "/home/ws/CoodWorkRun/FENet-v1/main.py", line 35, in mainrunner = Runner(cfg)File "/home/ws/CoodWorkRun/FENet-v1/fenet/engine/runner.py", line 29, in __init__self.recorder = build_recorder(self.cfg)File "/home/ws/CoodWorkRun/FENet-v1/fenet/utils/recorder.py", line 135, in build_recorderreturn Recorder(cfg)File "/home/ws/CoodWorkRun/FENet-v1/fenet/utils/recorder.py", line 51, in __init__self.cp_projects(self.work_dir)File "/home/ws/CoodWorkRun/FENet-v1/fenet/utils/recorder.py", line 68, in cp_projectswith open('./.gitignore', 'r') as fp:
FileNotFoundError: [Errno 2] No such file or directory: './.gitignore'
解决办法：
# 1、使用cd命令更改当前工作目录到脚本所在的目录：
cd /home/ws/CoodWorkRun/FENet-v1
# 2、确认您已经在正确的目录中，可以使用pwd命令查看当前工作目录：
pwd
问题4：
指令：
(fenetv2) ws@sun:~$ python /home/ws/CoodWorkRun/FENet-v1/main.py
报错
AttributeError: module 'numpy' has no attribute 'bool'.
`np.bool` was a deprecated alias for the builtin `bool`. To avoid this error in existing code, use `bool` by itself. Doing this will not modify any behavior and is safe. If you specifically wanted the numpy scalar type, use `np.bool_` here.
The aliases was originally deprecated in NumPy 1.20; for more details and guidance see the original release note at:https://numpy.org/devdocs/release/1.20.0-notes.html#deprecations
解决办法：
# 这个问题是由于numpy的版本问题
# 1.22或者1.24都容易出现这个问题
# 所以我们需要将numpy换成1.23的版本
# 可以使用如下命令
pip install numpy==1.23.2
问题5：
指令：
(fenetv2) ws@sun:~$ python /home/ws/CoodWorkRun/FENet-v1/main.py
报错
The above exception was the direct cause of the following exception:Traceback (most recent call last):File "main.py", line 62, in <module>main()File "main.py", line 42, in mainrunner.train()File "/home/ws/CoodWorkRun/FENet-v1/fenet/engine/runner.py", line 128, in trainself.validate()File "/home/ws/CoodWorkRun/FENet-v1/fenet/engine/runner.py", line 173, in validatemetric = self.val_loader.dataset.evaluate(predictions,File "/home/ws/CoodWorkRun/FENet-v1/fenet/datasets/culane.py", line 157, in evaluateresult = culane_metric.eval_predictions(output_basedir,File "/home/ws/CoodWorkRun/FENet-v1/fenet/utils/culane_metric.py", line 243, in eval_predictionsresults = p.starmap(culane_metric, zip(predictions, annotations,File "/home/ws/anaconda3/envs/fenetv2/lib/python3.8/multiprocessing/pool.py", line 372, in starmapreturn self._map_async(func, iterable, starmapstar, chunksize).get()File "/home/ws/anaconda3/envs/fenetv2/lib/python3.8/multiprocessing/pool.py", line 771, in getraise self._value
ValueError: Invalid inputs.
解决办法：
# fenet/utils/culane_metric.py 路径下interp函数修改如下：
def interp(points, n=50):if len(points) < 2: # 至少需要2个点进行插值print("Not enough points for interpolation.")return np.array(points)x = [p[0] for p in points]y = [p[1] for p in points]try:# 尝试进行样条曲线拟合tck, u = splprep([x, y], s=0, k=min(3, len(points) - 1))u_new = np.linspace(0, 1, num=(len(u) - 1) * n + 1)out = splev(u_new, tck)return np.array(out).T # 将输出转换为适当的形式except Exception as e:print(f"Error in spline interpolation: {e}")return np.array(points) # 在错误情况下返回原始点
