conv2d() got an unexpected keyword argument 'border_mode' about sam

[marcellacornia]

Hi @Mageshwaran2314,
thanks for downloading our code.

Please check your Keras and Theano versions. Our code is compatible with Keras 1.1.0 and Theano 0.9.0.

[Mageshwaran2314]

yes I try to run this code with keras 2.1.3

I made some changes to the code, after that, I got this error

Using Theano backend.
Traceback (most recent call last):
File "main.py", line 63, in
m = Model(input=[x, x_maps], output=sam_resnet([x, x_maps]))
File "E:\sam-master\models.py", line 130, in sam_resnet
dcn = dcn_resnet(input_tensor=x[0])
File "E:\sam-master\dcn_resnet.py", line 143, in dcn_resnet
x = conv_block(x, 3, [64, 64, 256], stage=2, block='a', strides=(1, 1))
File "E:\sam-master\dcn_resnet.py", line 50, in conv_block
name=conv_name_base + '2a')(input_tensor)
File "C:\ProgramData\Anaconda3\lib\site-packages\keras\legacy\interfaces.py", line 56, in wrapper
raise_duplicate_arg_error(old_name, new_name)
File "C:\ProgramData\Anaconda3\lib\site-packages\keras\legacy\interfaces.py", line 106, in raise_duplicate_arg_error
'' + new_arg + '. Stick to the latter!')
TypeError: For the strides argument, the layer received both the legacy keyword argument subsample and the Keras 2 keyword argument strides. Stick to the latter!

Help me with this

[marcellacornia]

It seems that you are still using Keras 2.

You have to check your keras.json file. It should be in the following format:

{
    "image_dim_ordering": "th", 
    "epsilon": 1e-07, 
    "floatx": "float32", 
    "backend": "theano"
}

[Mageshwaran2314]

already I change this
{
"floatx": "float32",
"epsilon": 1e-07,
"image_dim_ordering": "th",
"backend": "theano"
}

Traceback (most recent call last):
File "/home/bl/PycharmProjects/Predicting Human Eye Fixations via an LSTM-based Saliency Attentive Model/main.py", line 66, in
m = Model(input=[x, x_maps], output=sam_vgg([x, x_maps]))
File "/home/bl/PycharmProjects/Predicting Human Eye Fixations via an LSTM-based Saliency Attentive Model/models.py", line 107, in sam_vgg
nb_cols=3, nb_rows=3)(att_convlstm)
File "/usr/local/lib/python2.7/dist-packages/Keras-2.0.3-py2.7.egg/keras/engine/topology.py", line 578, in call
output = self.call(inputs, **kwargs)
File "/home/bl/PycharmProjects/Predicting Human Eye Fixations via an LSTM-based Saliency Attentive Model/attentive_convlstm.py", line 143, in call
initial_states = self.get_initial_states(x)
File "/home/bl/PycharmProjects/Predicting Human Eye Fixations via an LSTM-based Saliency Attentive Model/attentive_convlstm.py", line 42, in get_initial_states
initial_state = K.conv2d(initial_state, K.zeros((self.nb_filters_out, self.nb_filters_in, 1, 1)), border_mode='same')
TypeError: conv2d() got an unexpected keyword argument 'border_mode'
why?

import keras
print keras.version
import theano as th
print th.version

2.0.3
0.8.2

Traceback (most recent call last):
File "/home/bl/PycharmProjects/Predicting Human Eye Fixations via an LSTM-based Saliency Attentive Model/main.py", line 77, in
m = Model(input=[x, x_maps], output=sam_resnet([x, x_maps]))
File "/home/bl/PycharmProjects/Predicting Human Eye Fixations via an LSTM-based Saliency Attentive Model/models.py", line 136, in sam_resnet
nb_cols=3, nb_rows=3)(att_convlstm)
File "/usr/local/lib/python2.7/dist-packages/Keras-2.0.3-py2.7.egg/keras/engine/topology.py", line 578, in call
output = self.call(inputs, **kwargs)
File "/home/bl/PycharmProjects/Predicting Human Eye Fixations via an LSTM-based Saliency Attentive Model/attentive_convlstm.py", line 147, in call
initial_states = self.get_initial_states(x)
File "/home/bl/PycharmProjects/Predicting Human Eye Fixations via an LSTM-based Saliency Attentive Model/attentive_convlstm.py", line 43, in get_initial_states
initial_state = K.conv2d(initial_state, K.zeros((self.nb_filters_out, self.nb_filters_in, 1, 1)), padding='same')
File "/usr/local/lib/python2.7/dist-packages/Keras-2.0.3-py2.7.egg/keras/backend/tensorflow_backend.py", line 2921, in conv2d
data_format='NHWC')
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/nn_ops.py", line 650, in convolution
num_spatial_dims]))
ValueError: number of input channels does not match corresponding dimension of filter, 512 != 1

[ZhenyueQin]

initial_state = K.conv2d(initial_state, K.zeros((self.nb_filters_out, self.nb_filters_in, 1, 1)), padding='same')
This solves the issue