Python 如何将具有自定义keras图层(.h5)的keras模型冻结为tensorflow图(.pb)?
我正在尝试实现一个更快的目标检测RCNN模型。在我使用Python 如何将具有自定义keras图层(.h5)的keras模型冻结为tensorflow图(.pb)?,python,tensorflow,keras,Python,Tensorflow,Keras,我正在尝试实现一个更快的目标检测RCNN模型。在我使用model\u all.save('filename.h5')训练并保存模型后,我尝试将Keras模型冻结为TensorFlow图(as.pb),以便使用Amir Abdi编写的进行推理。但是当我尝试转换它时,由于自定义的roipoolgconv层,我得到了一个ValueError:Unknown层:roipoolgconv: class RoiPoolingConv(Layer): '''ROI pooling layer for 2D i
model\u all.save('filename.h5')
训练并保存模型后,我尝试将Keras模型冻结为TensorFlow图(as.pb
),以便使用Amir Abdi编写的进行推理。但是当我尝试转换它时,由于自定义的roipoolgconv
层,我得到了一个ValueError:Unknown层:roipoolgconv
:
class RoiPoolingConv(Layer):
'''ROI pooling layer for 2D inputs.
See Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition,
K. He, X. Zhang, S. Ren, J. Sun
# Arguments
pool_size: int
Size of pooling region to use. pool_size = 7 will result in a 7x7 region.
num_rois: number of regions of interest to be used
# Input shape
list of two 4D tensors [X_img,X_roi] with shape:
X_img:
`(1, rows, cols, channels)`
X_roi:
`(1,num_rois,4)` list of rois, with ordering (x,y,w,h)
# Output shape
3D tensor with shape:
`(1, num_rois, channels, pool_size, pool_size)`
'''
def __init__(self, pool_size, num_rois, **kwargs):
self.dim_ordering = K.image_dim_ordering()
self.pool_size = pool_size
self.num_rois = num_rois
super(RoiPoolingConv, self).__init__(**kwargs)
def build(self, input_shape):
self.nb_channels = input_shape[0][3]
def compute_output_shape(self, input_shape):
return None, self.num_rois, self.pool_size, self.pool_size, self.nb_channels
def call(self, x, mask=None):
assert(len(x) == 2)
# x[0] is image with shape (rows, cols, channels)
img = x[0]
# x[1] is roi with shape (num_rois,4) with ordering (x,y,w,h)
rois = x[1]
input_shape = K.shape(img)
outputs = []
for roi_idx in range(self.num_rois):
x = rois[0, roi_idx, 0]
y = rois[0, roi_idx, 1]
w = rois[0, roi_idx, 2]
h = rois[0, roi_idx, 3]
x = K.cast(x, 'int32')
y = K.cast(y, 'int32')
w = K.cast(w, 'int32')
h = K.cast(h, 'int32')
# Resized roi of the image to pooling size (7x7)
rs = tf.image.resize_images(img[:, y:y+h, x:x+w, :], (self.pool_size, self.pool_size))
outputs.append(rs)
final_output = K.concatenate(outputs, axis=0)
# Reshape to (1, num_rois, pool_size, pool_size, nb_channels)
# Might be (1, 4, 7, 7, 3)
final_output = K.reshape(final_output, (1, self.num_rois, self.pool_size, self.pool_size, self.nb_channels))
# permute_dimensions is similar to transpose
final_output = K.permute_dimensions(final_output, (0, 1, 2, 3, 4))
return final_output
def get_config(self):
config = {'pool_size': self.pool_size,
'num_rois': self.num_rois}
base_config = super(RoiPoolingConv, self).get_config()
return dict(list(base_config.items()) + list(config.items()))
我已经查看了大部分资源,几乎所有资源都建议对这一层进行注释。但是,由于这一层对于目标检测很重要,我想知道是否有一个解决办法
错误的完整回溯(注意:我已将文件名保存为freezekeras.py
,内容与keras\u to\u tensorflow.py
)相同:
使用TensorFlow后端。
回溯(最近一次呼叫最后一次):
文件“freezekeras.py”,第181行,在
应用程序运行(主)
文件“/usr/local/lib/python3.5/dist-packages/absl/app.py”,第300行,运行中
_运行_main(main,args)
文件“/usr/local/lib/python3.5/dist-packages/absl/app.py”,第251行,在主
系统出口(主(argv))
文件“freezekeras.py”,第127行,在main中
model=load\u model(FLAGS.input\u model、FLAGS.input\u model\u json、FLAGS.input\u model\u yaml)
文件“freezekeras.py”,第105行,在load_模型中
提出错误的\u文件\u错误
文件“freezekeras.py”,第62行,在load_模型中
模型=keras.models.load\u模型(输入模型路径)
文件“/usr/local/lib/python3.5/dist-packages/keras/engine/saving.py”,第419行,在load_模型中
model=\反序列化\模型(f,自定义\对象,编译)
文件“/usr/local/lib/python3.5/dist-packages/keras/engine/saving.py”,第225行,在反序列化模型中
模型=来自配置的模型(模型配置,自定义对象=自定义对象)
文件“/usr/local/lib/python3.5/dist-packages/keras/engine/saving.py”,第458行,型号为
返回反序列化(配置,自定义对象=自定义对象)
文件“/usr/local/lib/python3.5/dist-packages/keras/layers/_-init__.py”,第55行,反序列化
可打印\u模块\u name='layer')
文件“/usr/local/lib/python3.5/dist packages/keras/utils/generic_utils.py”,第145行,反序列化_keras_对象
列表(自定义对象.项())
文件“/usr/local/lib/python3.5/dist packages/keras/engine/network.py”,第1022行,from_config
处理层(层数据)
文件“/usr/local/lib/python3.5/dist-packages/keras/engine/network.py”,第1008行,进程层中
自定义对象=自定义对象)
文件“/usr/local/lib/python3.5/dist-packages/keras/layers/_-init__.py”,第55行,反序列化
可打印\u模块\u name='layer')
文件“/usr/local/lib/python3.5/dist packages/keras/utils/generic_utils.py”,第138行,反序列化_keras_对象
“:”+类名)
ValueError:未知层:ROIPoolGCONV
尝试明确指定自定义图层:
model = load_model('my_model.h5', custom_objects={'RoiPoolingConv': RoiPoolingConv})
显然,您必须将keras\u重新编写为\u tensorflow.py
脚本。请参见解决方案下的“保存的模型”部分中的“处理自定义图层(或其他自定义对象)”
keras\u to\u tensorflow.py中加载模型时指定自定义层
请提供错误的完整跟踪堆栈。确定!我加了,谢谢!我没有在正确的位置进行搜索,我修改了代码,它工作正常。
model = load_model('my_model.h5', custom_objects={'RoiPoolingConv': RoiPoolingConv})
model = keras.models.load_model(input_model_path, custom_objects={'RoiPoolingConv':RoiPoolingConv})
def __init__(self, pool_size = 7, num_rois = 32, **kwargs):