我尝试使用 Tensorflow 2.15 进行一个带有对象检测的小项目，如下所示：https://github.com/nicknochnack/TFODCourse。当我尝试安装用于对象检测和协议的模块时，我发现了一些错误。
在使用 model_builder_tf2_test.py 测试验证脚本时遇到：
2024-03-19 20:38:37.057863: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
2024-03-19 20:38:37.802548: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
Traceback (most recent call last):
File "D:\TensorFlow Object Detection\TFODCourse\Tensorflow\models\research\object_detection\builders\model_builder_tf2_test.py", line 24, in <module>
from object_detection.builders import model_builder
File "D:\TensorFlow Object Detection\TFODCourse\tfod\Lib\site-packages\object_detection-0.1-py3.11.egg\object_detection\builders\model_builder.py", line 26, in <module>
from object_detection.builders import hyperparams_builder
File "D:\TensorFlow Object Detection\TFODCourse\tfod\Lib\site-packages\object_detection-0.1-py3.11.egg\object_detection\builders\hyperparams_builder.py", line 27, in <module>
from object_detection.core import freezable_sync_batch_norm
File "D:\TensorFlow Object Detection\TFODCourse\tfod\Lib\site-packages\object_detection-0.1-py3.11.egg\object_detection\core\freezable_sync_batch_norm.py", line 20, in <module>
class FreezableSyncBatchNorm(tf.keras.layers.experimental.SyncBatchNormalization
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
AttributeError: module 'keras._tf_keras.keras.layers' has no attribute 'experimental'
这是安装tensorflow对象检测和协议的代码：
import os
CUSTOM_MODEL_NAME = 'my_ssd_mobnet'
PRETRAINED_MODEL_NAME = 'ssd_mobilenet_v2_fpnlite_320x320_coco17_tpu-8'
PRETRAINED_MODEL_URL = 'http://download.tensorflow.org/models/object_detection/tf2/20200711/ssd_mobilenet_v2_fpnlite_320x320_coco17_tpu-8.tar.gz'
TF_RECORD_SCRIPT_NAME = 'generate_tfrecord.py'
LABEL_MAP_NAME = 'label_map.pbtxt'
paths = {
'WORKSPACE_PATH': os.path.join('Tensorflow', 'workspace'),
'SCRIPTS_PATH': os.path.join('Tensorflow','scripts'),
'APIMODEL_PATH': os.path.join('Tensorflow','models'),
'ANNOTATION_PATH': os.path.join('Tensorflow', 'workspace','annotations'),
'IMAGE_PATH': os.path.join('Tensorflow', 'workspace','images'),
'MODEL_PATH': os.path.join('Tensorflow', 'workspace','models'),
'PRETRAINED_MODEL_PATH': os.path.join('Tensorflow', 'workspace','pre-trained-models'),
'CHECKPOINT_PATH': os.path.join('Tensorflow', 'workspace','models',CUSTOM_MODEL_NAME),
'OUTPUT_PATH': os.path.join('Tensorflow', 'workspace','models',CUSTOM_MODEL_NAME, 'export'),
'TFJS_PATH':os.path.join('Tensorflow', 'workspace','models',CUSTOM_MODEL_NAME, 'tfjsexport'),
'TFLITE_PATH':os.path.join('Tensorflow', 'workspace','models',CUSTOM_MODEL_NAME, 'tfliteexport'),
'PROTOC_PATH':os.path.join('Tensorflow','protoc')
}
files = {
'PIPELINE_CONFIG':os.path.join('Tensorflow', 'workspace','models', CUSTOM_MODEL_NAME, 'pipeline.config'),
'TF_RECORD_SCRIPT': os.path.join(paths['SCRIPTS_PATH'], TF_RECORD_SCRIPT_NAME),
'LABELMAP': os.path.join(paths['ANNOTATION_PATH'], LABEL_MAP_NAME)
}
for path in paths.values():
if not os.path.exists(path):
if os.name == 'posix':
!mkdir -p {path}
if os.name == 'nt':
!mkdir {path}
if os.name=='nt':
!pip install wget
import wget
if not os.path.exists(os.path.join(paths['APIMODEL_PATH'], 'research', 'object_detection')):
!git clone https://github.com/tensorflow/models {paths['APIMODEL_PATH']}
# Install Tensorflow Object Detection
if os.name=='posix':
!apt-get install protobuf-compiler
!cd Tensorflow/models/research && protoc object_detection/protos/*.proto --python_out=. && cp object_detection/packages/tf2/setup.py . && python -m pip install .
if os.name=='nt':
url="https://github.com/protocolbuffers/protobuf/releases/download/v26.0/protoc-26.0-win64.zip"
wget.download(url)
!move protoc-26.0-win64.zip {paths['PROTOC_PATH']}
!cd {paths['PROTOC_PATH']} && tar -xf protoc-26.0-win64.zip
os.environ['PATH'] += os.pathsep + os.path.abspath(os.path.join(paths['PROTOC_PATH'], 'bin'))
!cd Tensorflow/models/research && protoc object_detection/protos/*.proto --python_out=. && copy object_detection\\packages\\tf2\\setup.py setup.py && python setup.py build && python setup.py install
!cd Tensorflow/models/research/slim && pip install -e .
VERIFICATION_SCRIPT = os.path.join(paths['APIMODEL_PATH'], 'research', 'object_detection', 'builders', 'model_builder_tf2_test.py')
# Verify Installation
!python {VERIFICATION_SCRIPT}
我该如何解决这个问题？我提到我有最新的tensorflow和pip库，并从以下位置克隆了tensorflow：https://github.com/tensorflow/models，Python版本为3.11。
我一直在尝试更新所有隐含的库，并在网上找到了这个：https://www.tensorflow.org/api_docs/python/tf/keras/layers/experimental/SyncBatchNormalization。即使它已被弃用，它们还没有在官方目录中更新吗？：https://github.com/tensorflow/models/tree/master/research/object_detection
嗯，这部分很有趣。
if os.name=='nt':
url="https://github.com/protocolbuffers/protobuf/releases/download/v26.0/protoc-26.0-win64.zip"
wget.download(url)
!move protoc-26.0-win64.zip {paths['PROTOC_PATH']}
!cd {paths['PROTOC_PATH']} && tar -xf protoc-26.0-win64.zip
os.environ['PATH'] += os.pathsep + os.path.abspath(os.path.join(paths['PROTOC_PATH'], 'bin'))
!cd Tensorflow/models/research && protoc object_detection/protos/*.proto --python_out=. && copy object_detection\\packages\\tf2\\setup.py setup.py && python setup.py build && python setup.py install
!cd Tensorflow/models/research/slim && pip install -e .
虽然这不是我安装软件包的传统方式，但它似乎依赖于特定版本，因为 PYPI 版本 相当旧（2019 年）。您的安装似乎是最好的方法。
我想有几条路可以走：
class FreezableSyncBatchNorm(tf.keras.layers.experimental.SyncBatchNormalization
->
class FreezableSyncBatchNorm(tf.keras.layers.BatchNormalization
如here所述，使用 BatchNormalization 并传递
synchronized=True