ImportError:没有名为'nets'的模块

Question

我试图从训练集/流程模型中提供的export_inference_graph.py脚本中将trained_checkpoint转换为最终冻结模型,但会出现以下错误结果.是的,我已经设置了$ PYTHONPATH"模特/苗条",但我仍然得到这个错误,有人可以帮助我吗？

$ echo $PYTHONPATH
:/home/ishara/tensorflow_models/models:/home/ishara/tensorflow_models/models/slim

*****************************问题****************************************************************************

$sudo python3 object_detection/export_inference_graph.py  --input_type image_tensor  --pipeline_config_path = "ssd_inception_v2_pets.config"  --trained_checkpoint_prefix="output/model.ckpt-78543"  --output_directory="birds_inference_graph.pb"

Traceback (most recent call last):
  File "object_detection/export_inference_graph.py", line 74, in <module>
    from object_detection import exporter
  File "/usr/local/lib/python3.5/dist-packages/object_detection-0.1-py3.5.egg/object_detection/exporter.py", line 28, in <module>

  File "/usr/local/lib/python3.5/dist-packages/object_detection-0.1-py3.5.egg/object_detection/builders/model_builder.py", line 30, in <module>
  File "/usr/local/lib/python3.5/dist-packages/object_detection-0.1-py3.5.egg/object_detection/models/faster_rcnn_inception_resnet_v2_feature_extractor.py", line 28, in <module>
ImportError: No module named 'nets'

---

我已经苦苦挣扎了好几天了,试过很多解决方案没什么用,我正在使用带有tensorflow-gpu版本的Ubuntu 16.04.

Answer 1

在https://github.com/tensorflow/models/blob/master/research/object_detection/g3doc/installation.md查看Protobuf编译 并正确设置PYTHONPATH,这就是我为Windows解决这个问题的方法

对于Windows:

来自tensorflow/models/research /

步骤1: protoc object_detection/protos/*.proto --python_out=.

第2步:

set PYTHONPATH= <Path to 'research' Directory> ; <Path to 'slim' Directory>

对于Eg:

set PYTHONPATH=C:\Users\Guy\Desktop\models\research;C:\Users\Guy\Desktop\models\research\slim

Answer 2

Ubuntu 18 物理 GPU（设备：0，名称：Tesla K80，pci 总线 ID：0000:00:1e.0，计算能力：3.7）强文本

cd models/research/slim/;
python setup.py build
python setup.py install

以防万一您已经下载或克隆了您的模型目录。

Answer 3

TF-Slim 可通过 TensorFlow 1.0 以 tf.contrib.slim 形式提供，因此如果您使用pip install tensorflow. 你还需要做这三件事：

1. 安装模型库

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

2. 添加PYTHONPATH到.bashrc

   $ cd
   $ vi .bashrc
   
   export PYTHONPATH="$PYTHONPATH:/home/${YOUR_USERNAME}/models/research/slim"
   

3. 将模型路径添加到脚本中

   $ vi ${YOUR_SCRIPT}.py
   
   import sys
   sys.path.append('/home/${YOUR_USERNAME}/models/research/slim/')
   

完成这 3 个步骤后，一切就都准备好了。现在您可以像这样导入 TF 网络：

import tensorflow as tf
from tensorflow.contrib import slim
from nets import inception_resnet_v2