ValueError: Output tensors to a Model must be the output of a TensorFlow Layer (thus holding past layer metadata). Found: Tensor("mrcnn_class_logits_5/transpose_1:0", shape=(4, ?, 3), dtype=float32)

[ruchika61]

No description provided.

[ruchika61]

hi
i am getting this error at
train_maskrcnn.load_pretrained_model("mask_rcnn_coco.h5")
kindly tell the soluion.
Also for custom dataset i am getting error
RuntimeError: Exception thrown in SimpleITK ReadImage: D:\a\1\sitk\Code\IO\src\sitkImageReaderBase.cxx:107:
sitk::ERROR: Unable to determine ImageIO reader for "C:\Users\dell\Desktop\Custom Semantic Segmentation\chest\train\17.png"
at
vis_img.visualize_sample().
waiting for your reply.

[ruchika61]

hi
i am getting above error at
train_maskrcnn.load_pretrained_model("mask_rcnn_coco.h5")
kindly tell the soluion.
Also for custom dataset i am getting error
RuntimeError: Exception thrown in SimpleITK ReadImage: D:\a\1\sitk\Code\IO\src\sitkImageReaderBase.cxx:107:
sitk::ERROR: Unable to determine ImageIO reader for "C:\Users\dell\Desktop\Custom Semantic Segmentation\chest\train\17.png"
at
vis_img.visualize_sample().
waiting for your reply.

at
train_maskrcnn.load_pretrained_model("mask_rcnn_coco.h5")
kindly tell the soluion.
Also for custom dataset i am getting error
RuntimeError: Exception thrown in SimpleITK ReadImage: D:\a\1\sitk\Code\IO\src\sitkImageReaderBase.cxx:107:
sitk::ERROR: Unable to determine ImageIO reader for "C:\Users\dell\Desktop\Custom Semantic Segmentation\chest\train\17.png"
at
vis_img.visualize_sample().
waiting for your reply.

[ayoolaolafenwa]

Provide your full code. Do you prepare the dataset with labelme? Does the number of classes you pass into the code match the classes available in the dataset?

[ruchika61]

Hi
Thanks for your reply
Yes, I have prepared dataset to labelme. Yes, I have entered same number of classes into code.
code is same as yours
I tried first to work with Nature dataset as you said. But in that i am getting this error

runfile('C:/Users/dell/Desktop/Custom Semantic Segmentation/demo.py', wdir='C:/Users/dell/Desktop/Custom Semantic Segmentation')
Reloaded modules: SimpleITK, SimpleITK.SimpleITK, SimpleITK._SimpleITK
Using resnet101 as network backbone For Mask R-CNN model
WARNING:tensorflow:From C:\Users\dell\Anaconda3\lib\site-packages\tensorflow\python\ops\resource_variable_ops.py:435: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.
Instructions for updating:
Colocations handled automatically by placer.
Traceback (most recent call last):

File "", line 1, in
runfile('C:/Users/dell/Desktop/Custom Semantic Segmentation/demo.py', wdir='C:/Users/dell/Desktop/Custom Semantic Segmentation')

File "C:\Users\dell\Anaconda3\lib\site-packages\spyder_kernels\customize\spydercustomize.py", line 704, in runfile
execfile(filename, namespace)

File "C:\Users\dell\Anaconda3\lib\site-packages\spyder_kernels\customize\spydercustomize.py", line 108, in execfile
exec(compile(f.read(), filename, 'exec'), namespace)

File "C:/Users/dell/Desktop/Custom Semantic Segmentation/demo.py", line 15, in
train_maskrcnn.load_pretrained_model("mask_rcnn_coco.h5")

File "C:\Users\dell\Anaconda3\lib\site-packages\pixellib\custom_train.py", line 50, in load_pretrained_model
self.model = modellib.MaskRCNN(mode="training", model_dir = self.model_dir, config=self.config)

File "C:\Users\dell\Anaconda3\lib\site-packages\pixellib\mask_rcnn.py", line 1822, in init
self.keras_model = self.build(mode=mode, config=config)

File "C:\Users\dell\Anaconda3\lib\site-packages\pixellib\mask_rcnn.py", line 2017, in build
model = KM.Model(inputs, outputs, name='mask_rcnn')

File "C:\Users\dell\Anaconda3\lib\site-packages\tensorflow\python\keras\engine\training.py", line 121, in init
super(Model, self).init(*args, **kwargs)

File "C:\Users\dell\Anaconda3\lib\site-packages\tensorflow\python\keras\engine\network.py", line 81, in init
self._init_graph_network(*args, **kwargs)

File "C:\Users\dell\Anaconda3\lib\site-packages\tensorflow\python\training\checkpointable\base.py", line 442, in _method_wrapper
method(self, *args, **kwargs)

File "C:\Users\dell\Anaconda3\lib\site-packages\tensorflow\python\keras\engine\network.py", line 177, in _init_graph_network
self._validate_graph_inputs_and_outputs()

File "C:\Users\dell\Anaconda3\lib\site-packages\tensorflow\python\keras\engine\network.py", line 1673, in _validate_graph_inputs_and_outputs
'(thus holding past layer metadata). Found: ' + str(x))

ValueError: Output tensors to a Model must be the output of a TensorFlow Layer (thus holding past layer metadata). Found: Tensor("mrcnn_class_logits/transpose_1:0", shape=(4, ?, 3), dtype=float32)

[ruchika61]

when i tried on my custom dataset and chenged load_dataset function Nature with custom dataset, i am getting this error

runfile('C:/Users/dell/Desktop/Custom Semantic Segmentation/chest.py', wdir='C:/Users/dell/Desktop/Custom Semantic Segmentation')
Traceback (most recent call last):

File "", line 1, in
runfile('C:/Users/dell/Desktop/Custom Semantic Segmentation/chest.py', wdir='C:/Users/dell/Desktop/Custom Semantic Segmentation')

File "C:\Users\dell\Anaconda3\lib\site-packages\spyder_kernels\customize\spydercustomize.py", line 704, in runfile
execfile(filename, namespace)

File "C:\Users\dell\Anaconda3\lib\site-packages\spyder_kernels\customize\spydercustomize.py", line 108, in execfile
exec(compile(f.read(), filename, 'exec'), namespace)

File "C:/Users/dell/Desktop/Custom Semantic Segmentation/chest.py", line 12, in
vis_img.visualize_sample()

File "C:\Users\dell\Anaconda3\lib\site-packages\pixellib\custom_train.py", line 88, in visualize_sample
image = self.dataset_train.load_image(image_id)

File "C:\Users\dell\Anaconda3\lib\site-packages\pixellib\utils.py", line 362, in load_image
image = skimage.io.imread(self.image_info[image_id]['path'])

File "C:\Users\dell\Anaconda3\lib\site-packages\skimage\io_io.py", line 61, in imread
img = call_plugin('imread', fname, plugin=plugin, **plugin_args)

File "C:\Users\dell\Anaconda3\lib\site-packages\skimage\io\manage_plugins.py", line 210, in call_plugin
return func(*args, **kwargs)

File "C:\Users\dell\Anaconda3\lib\site-packages\imageio\core\functions.py", line 221, in imread
reader = read(uri, format, "i", **kwargs)

File "C:\Users\dell\Anaconda3\lib\site-packages\imageio\core\functions.py", line 143, in get_reader
return format.get_reader(request)

File "C:\Users\dell\Anaconda3\lib\site-packages\imageio\core\format.py", line 174, in get_reader
return self.Reader(self, request)

File "C:\Users\dell\Anaconda3\lib\site-packages\imageio\core\format.py", line 224, in init
self._open(**self.request.kwargs.copy())

File "C:\Users\dell\Anaconda3\lib\site-packages\imageio\plugins\simpleitk.py", line 84, in _open
self._img = _itk.ReadImage(self.request.get_local_filename())

File "C:\Users\dell\AppData\Roaming\Python\Python37\site-packages\SimpleITK\SimpleITK.py", line 8876, in ReadImage
return _SimpleITK.ReadImage(*args)

RuntimeError: Exception thrown in SimpleITK ReadImage: D:\a\1\sitk\Code\IO\src\sitkImageReaderBase.cxx:107:
sitk::ERROR: Unable to determine ImageIO reader for "C:\Users\dell\Desktop\Custom Semantic Segmentation\chest\train\97.png"

Codes are common and same as:
import pixellib
from pixellib.custom_train import instance_custom_training

vis_img = instance_custom_training()
vis_img.load_dataset("chest")
vis_img.visualize_sample()

train_maskrcnn.modelConfig(network_backbone = "resnet101", num_classes= 7, batch_size = 4)
train_maskrcnn.load_pretrained_model("mask_rcnn_coco.h5")
train_maskrcnn.load_dataset("chest")
train_maskrcnn.train_model(num_epochs = 300, augmentation=True, path_trained_models = "mask_rcnn_models")

[ayoolaolafenwa]

Hi
Thanks for your reply
Yes, I have prepared dataset to labelme. Yes, I have entered same number of classes into code.
code is same as yours
I tried first to work with Nature dataset as you said. But in that i am getting this error

runfile('C:/Users/dell/Desktop/Custom Semantic Segmentation/demo.py', wdir='C:/Users/dell/Desktop/Custom Semantic Segmentation')
Reloaded modules: SimpleITK, SimpleITK.SimpleITK, SimpleITK._SimpleITK
Using resnet101 as network backbone For Mask R-CNN model
WARNING:tensorflow:From C:\Users\dell\Anaconda3\lib\site-packages\tensorflow\python\ops\resource_variable_ops.py:435: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.
Instructions for updating:
Colocations handled automatically by placer.
Traceback (most recent call last):

File "", line 1, in
runfile('C:/Users/dell/Desktop/Custom Semantic Segmentation/demo.py', wdir='C:/Users/dell/Desktop/Custom Semantic Segmentation')

File "C:\Users\dell\Anaconda3\lib\site-packages\spyder_kernels\customize\spydercustomize.py", line 704, in runfile
execfile(filename, namespace)

File "C:\Users\dell\Anaconda3\lib\site-packages\spyder_kernels\customize\spydercustomize.py", line 108, in execfile
exec(compile(f.read(), filename, 'exec'), namespace)

File "C:/Users/dell/Desktop/Custom Semantic Segmentation/demo.py", line 15, in
train_maskrcnn.load_pretrained_model("mask_rcnn_coco.h5")

File "C:\Users\dell\Anaconda3\lib\site-packages\pixellib\custom_train.py", line 50, in load_pretrained_model
self.model = modellib.MaskRCNN(mode="training", model_dir = self.model_dir, config=self.config)

File "C:\Users\dell\Anaconda3\lib\site-packages\pixellib\mask_rcnn.py", line 1822, in init
self.keras_model = self.build(mode=mode, config=config)

File "C:\Users\dell\Anaconda3\lib\site-packages\pixellib\mask_rcnn.py", line 2017, in build
model = KM.Model(inputs, outputs, name='mask_rcnn')

File "C:\Users\dell\Anaconda3\lib\site-packages\tensorflow\python\keras\engine\training.py", line 121, in init
super(Model, self).init(*args, **kwargs)

File "C:\Users\dell\Anaconda3\lib\site-packages\tensorflow\python\keras\engine\network.py", line 81, in init
self._init_graph_network(*args, **kwargs)

File "C:\Users\dell\Anaconda3\lib\site-packages\tensorflow\python\training\checkpointable\base.py", line 442, in _method_wrapper
method(self, *args, **kwargs)

File "C:\Users\dell\Anaconda3\lib\site-packages\tensorflow\python\keras\engine\network.py", line 177, in _init_graph_network
self._validate_graph_inputs_and_outputs()

File "C:\Users\dell\Anaconda3\lib\site-packages\tensorflow\python\keras\engine\network.py", line 1673, in _validate_graph_inputs_and_outputs
'(thus holding past layer metadata). Found: ' + str(x))

ValueError: Output tensors to a Model must be the output of a TensorFlow Layer (thus holding past layer metadata). Found: Tensor("mrcnn_class_logits/transpose_1:0", shape=(4, ?, 3), dtype=float32)

This is not an issue with PixelLib. Ensure that you have installed tensorflow properly. It appears like you are using anaconda3, ensure that you installed a compatible tensorflow version.

[ayoolaolafenwa]

when i tried on my custom dataset and chenged load_dataset function Nature with custom dataset, i am getting this error

runfile('C:/Users/dell/Desktop/Custom Semantic Segmentation/chest.py', wdir='C:/Users/dell/Desktop/Custom Semantic Segmentation')
Traceback (most recent call last):

File "", line 1, in
runfile('C:/Users/dell/Desktop/Custom Semantic Segmentation/chest.py', wdir='C:/Users/dell/Desktop/Custom Semantic Segmentation')

File "C:\Users\dell\Anaconda3\lib\site-packages\spyder_kernels\customize\spydercustomize.py", line 704, in runfile
execfile(filename, namespace)

File "C:\Users\dell\Anaconda3\lib\site-packages\spyder_kernels\customize\spydercustomize.py", line 108, in execfile
exec(compile(f.read(), filename, 'exec'), namespace)

File "C:/Users/dell/Desktop/Custom Semantic Segmentation/chest.py", line 12, in
vis_img.visualize_sample()

File "C:\Users\dell\Anaconda3\lib\site-packages\pixellib\custom_train.py", line 88, in visualize_sample
image = self.dataset_train.load_image(image_id)

File "C:\Users\dell\Anaconda3\lib\site-packages\pixellib\utils.py", line 362, in load_image
image = skimage.io.imread(self.image_info[image_id]['path'])

File "C:\Users\dell\Anaconda3\lib\site-packages\skimage\io_io.py", line 61, in imread
img = call_plugin('imread', fname, plugin=plugin, **plugin_args)

File "C:\Users\dell\Anaconda3\lib\site-packages\skimage\io\manage_plugins.py", line 210, in call_plugin
return func(*args, **kwargs)

File "C:\Users\dell\Anaconda3\lib\site-packages\imageio\core\functions.py", line 221, in imread
reader = read(uri, format, "i", **kwargs)

File "C:\Users\dell\Anaconda3\lib\site-packages\imageio\core\functions.py", line 143, in get_reader
return format.get_reader(request)

File "C:\Users\dell\Anaconda3\lib\site-packages\imageio\core\format.py", line 174, in get_reader
return self.Reader(self, request)

File "C:\Users\dell\Anaconda3\lib\site-packages\imageio\core\format.py", line 224, in init
self._open(**self.request.kwargs.copy())

File "C:\Users\dell\Anaconda3\lib\site-packages\imageio\plugins\simpleitk.py", line 84, in _open
self._img = _itk.ReadImage(self.request.get_local_filename())

File "C:\Users\dell\AppData\Roaming\Python\Python37\site-packages\SimpleITK\SimpleITK.py", line 8876, in ReadImage
return _SimpleITK.ReadImage(*args)

RuntimeError: Exception thrown in SimpleITK ReadImage: D:\a\1\sitk\Code\IO\src\sitkImageReaderBase.cxx:107:
sitk::ERROR: Unable to determine ImageIO reader for "C:\Users\dell\Desktop\Custom Semantic Segmentation\chest\train\97.png"

Codes are common and same as:
import pixellib
from pixellib.custom_train import instance_custom_training

vis_img = instance_custom_training()
vis_img.load_dataset("chest")
vis_img.visualize_sample()

train_maskrcnn.modelConfig(network_backbone = "resnet101", num_classes= 7, batch_size = 4)
train_maskrcnn.load_pretrained_model("mask_rcnn_coco.h5")
train_maskrcnn.load_dataset("chest")
train_maskrcnn.train_model(num_epochs = 300, augmentation=True, path_trained_models = "mask_rcnn_models")

This is also not an issue with PixelLib. Check these similar issues on Runtime Error on SimpleITK ReadImage, they may be helpful.

SimpleITK/SimpleITK#434

junqiangchen/KiTS19-Challege#4