我在使用 Tensorflow object_detection 的 python 项目中遇到此错误

Question

创建TF记录

代码：

!python {SCRIPTS_PATH + '/generate_tfrecord.py'} -x {IMAGE_PATH + '/train'} -l {ANNOTATION_PATH + '/label_map.pbtxt'} -o {ANNOTATION_PATH + '/train.record'}

!python {SCRIPTS_PATH + '/generate_tfrecord.py'} -x{IMAGE_PATH + '/test'} -l {ANNOTATION_PATH + '/label_map.pbtxt'} -o {ANNOTATION_PATH + '/test.record'}

错误：

Traceback (most recent call last):
File "Tensorflow/scripts/generate_tfrecord.py", line 29, in <module>
From object_detection.utils import dataset_util, label_map_util
ModuleNotFoundError: No module named 'object_detection'

Traceback (most recent call last):
File "Tensorflow/scripts/generate_tfrecord.py", line 29, in <module>
From object_detection.utils import dataset_util, label_map_util
ModuleNotFoundError: No module named 'object_detection'

我关注了尼古拉斯·雷诺特视频 https://youtu.be/pDXdlXlaCco@22:30

我从 git clone 下载了 Tensorflow object_detection 模型https://github.com/tensorflow/models

然后我就跑了

python -m install pip 

Answer 1

之后

git clone https://github.com/tensorflow/models

你必须运行 protoc 才能正确安装 API：

cd models/research/
protoc object_detection/protos/*.proto --python_out=.
cp object_detection/packages/tf2/setup.py .
python -m pip install .

毕竟用这个脚本测试它：

python object_detection/builders/model_builder_tf2_test.py

Answer 2

如果你已经安装了相关的依赖项，比如protoc，如下

%cd /content/models/research
!protoc object_detection/protos/*.proto --python_out=.

你需要通过以下方式创建环境，

!pip install tf_slim
pwd = os.getcwd()
os.environ['PYTHONPATH'] += f':{pwd}:{pwd}/slim' 

Answer 3

我遵循了@Jotunheim的建议

cd models/research/
protoc object_detection/protos/*.proto --python_out=.
cp object_detection/packages/tf2/setup.py .
python -m pip install .

尝试安装 PyYAML 5.4 时，第三个命令总是会失败

让它对我有用的是回到python3.9（3.10不起作用）并在运行第三行之前手动安装PyYAML 5.3.1

pip install PyYAML==5.3.1