def pool_eeg_signal_channel(input_eeg_list,
channel_idx,
concat_dim1=1,
is_rep=False):
'''
The function will use a bunch of image tensor and concat them with the multiplication of pooling size from the
channel dimension as concat_dim1=1. The arrangement should be every eeg height (channel) contains 5 kernel indexes
Args:
input_eeg_list: input image tensor list that obtained from split_image.py
channel_idx: the channel index based on different channel information
concat_dim1: concat dimension 1, the channel dimension
Returns: the concated kernel tensor and the tensor input shape for verification
'''
if not isinstance(input_eeg_list, list):
print("input eeg split signal is not a list of tensor.")
if len(input_eeg_list) > 0:
input_shape = input_eeg_list[0].get_shape().as_list()
else:
print("input eeg split not exist.")
return
# get the size of the vec1
# conv_idx = biosemi_chan_mapper.pool_mapper(channel_idx)
if is_rep: # this will only be used for pooling of resize channels
vec_idx = np.arange(0, EEG_SIGNAL_SIZE),
vec_idx = np.reshape(vec_idx, (EEG_SIGNAL_SIZE, 1))
conv_idx = np.repeat(vec_idx, BIOSEMI_POOL, axis=1)
else:
conv_idx = roi_mapper_idx.roi_mapper_idx_select(channel_idx, 'pool')
conv_idx = biosemi_chan_mapper.replace_mapper_idx(conv_idx)
conv_idx_shape = conv_idx.shape # (128, 2)
# input_shape = [10, 1, 256, 1]
input_shape[concat_dim1] *= conv_idx_shape[0]
input_shape[concat_dim1] *= conv_idx_shape[1]
# input_shape = [10, 2*128, 256, 1]
curr_kernel_tensor = []
for kernel_idx in conv_idx: # go for every kernel => 256 kernels
# for each kernel => 5 index
# concat on the first dimension => concat_dim1 = 1
curr_kernel_tensor.append(
tf.concat(concat_dim1,
[input_eeg_list[idx] for idx in kernel_idx]))
kernel_tensor = tf.concat(concat_dim1, curr_kernel_tensor)
return kernel_tensor, input_shape
def pool_eeg_signal_channel(
input_eeg_list,
channel_idx,
concat_dim1=1,
is_rep=False):
'''
The function will use a bunch of image tensor and concat them with the multiplication of pooling size from the
channel dimension as concat_dim1=1. The arrangement should be every eeg height (channel) contains 5 kernel indexes
Args:
input_eeg_list: input image tensor list that obtained from split_image.py
channel_idx: the channel index based on different channel information
concat_dim1: concat dimension 1, the channel dimension
Returns: the concated kernel tensor and the tensor input shape for verification
'''
if not isinstance(input_eeg_list, list):
print("input eeg split signal is not a list of tensor.")
if len(input_eeg_list) > 0:
input_shape = input_eeg_list[0].get_shape().as_list()
else:
print("input eeg split not exist.")
return
# get the size of the vec1
# conv_idx = biosemi_chan_mapper.pool_mapper(channel_idx)
if is_rep: # this will only be used for pooling of resize channels
vec_idx = np.arange(0, EEG_SIGNAL_SIZE),
vec_idx = np.reshape(vec_idx, (EEG_SIGNAL_SIZE, 1))
conv_idx = np.repeat(vec_idx, BIOSEMI_POOL, axis=1)
else:
conv_idx = roi_mapper_idx.roi_mapper_idx_select(channel_idx, 'pool')
conv_idx = biosemi_chan_mapper.replace_mapper_idx(conv_idx)
conv_idx_shape = conv_idx.shape # (128, 2)
# input_shape = [10, 1, 256, 1]
input_shape[concat_dim1] *= conv_idx_shape[0]
input_shape[concat_dim1] *= conv_idx_shape[1]
# input_shape = [10, 2*128, 256, 1]
curr_kernel_tensor = []
for kernel_idx in conv_idx: # go for every kernel => 256 kernels
# for each kernel => 5 index
# concat on the first dimension => concat_dim1 = 1
curr_kernel_tensor.append(tf.concat(concat_dim1, [input_eeg_list[idx] for idx in kernel_idx]))
kernel_tensor = tf.concat(concat_dim1, curr_kernel_tensor)
return kernel_tensor, input_shape
def conv_eeg_signal_channel(
input_eeg_list,
channel_idx,
concat_dim1=1):
'''
The function will use a bunch of image tensor and concat them with the multiplication of kernel size and from the
channel dimension as concat_dim1=1. The arrangement should be every eeg height (channel) contains 5 kernel indexes
Args:
input_eeg_list: input image tensor list that obtained from split_image.py
channel_idx: the channel index based on different channel information
concat_dim1: concat dimension 1, the channel dimension
Returns: the concated kernel tensor and the tensor input shape for verification
'''
if not isinstance(input_eeg_list, list):
print("input eeg split signal is not a list of tensor.")
if len(input_eeg_list) > 0:
input_shape = input_eeg_list[0].get_shape().as_list()
else:
print("input eeg split not exist.")
return
# get the size of the vec1
# conv_idx = biosemi_chan_mapper.conv_mapper(channel_idx)
conv_idx = roi_mapper_idx.roi_mapper_idx_select(channel_idx, 'conv')
conv_idx = biosemi_chan_mapper.replace_mapper_idx(conv_idx)
conv_idx_shape = conv_idx.shape # (256, 5)
# input_shape = [10, 1, 256, 1]
input_shape[concat_dim1] *= conv_idx_shape[0]
input_shape[concat_dim1] *= conv_idx_shape[1]
# input_shape = [10, 5*256, 256, 1]
curr_kernel_tensor = []
for kernel_idx in conv_idx: # go for every kernel => 256 kernels
# for each kernel => 5 index
# concat on the first dimension => concat_dim1 = 1
curr_kernel_tensor.append(tf.concat(concat_dim1, [input_eeg_list[idx] for idx in kernel_idx]))
kernel_tensor = tf.concat(concat_dim1, curr_kernel_tensor)
return kernel_tensor, input_shape
def conv_eeg_signal_channel(input_eeg_list, channel_idx, concat_dim1=1):
'''
The function will use a bunch of image tensor and concat them with the multiplication of kernel size and from the
channel dimension as concat_dim1=1. The arrangement should be every eeg height (channel) contains 5 kernel indexes
Args:
input_eeg_list: input image tensor list that obtained from split_image.py
channel_idx: the channel index based on different channel information
concat_dim1: concat dimension 1, the channel dimension
Returns: the concated kernel tensor and the tensor input shape for verification
'''
if not isinstance(input_eeg_list, list):
print("input eeg split signal is not a list of tensor.")
if len(input_eeg_list) > 0:
input_shape = input_eeg_list[0].get_shape().as_list()
else:
print("input eeg split not exist.")
return
# get the size of the vec1
# conv_idx = biosemi_chan_mapper.conv_mapper(channel_idx)
conv_idx = roi_mapper_idx.roi_mapper_idx_select(channel_idx, 'conv')
conv_idx = biosemi_chan_mapper.replace_mapper_idx(conv_idx)
conv_idx_shape = conv_idx.shape # (256, 5)
# input_shape = [10, 1, 256, 1]
input_shape[concat_dim1] *= conv_idx_shape[0]
input_shape[concat_dim1] *= conv_idx_shape[1]
# input_shape = [10, 5*256, 256, 1]
curr_kernel_tensor = []
for kernel_idx in conv_idx: # go for every kernel => 256 kernels
# for each kernel => 5 index
# concat on the first dimension => concat_dim1 = 1
curr_kernel_tensor.append(
tf.concat(concat_dim1,
[input_eeg_list[idx] for idx in kernel_idx]))
kernel_tensor = tf.concat(concat_dim1, curr_kernel_tensor)
return kernel_tensor, input_shape
def test_replace_mapper(self):
for chan_num in roi_property.CONV_CHAN_INFO:
conv_256_idx = biosemi_chan_mapper.conv_mapper(chan_num)
conv_256_idx = biosemi_chan_mapper.replace_mapper_idx(conv_256_idx)
self.assertEqual(conv_256_idx.shape, (chan_num, roi_property.BIOSEMI_CONV))
def test_replace_mapper(self):
for chan_num in roi_property.CONV_CHAN_INFO:
conv_256_idx = biosemi_chan_mapper.conv_mapper(chan_num)
conv_256_idx = biosemi_chan_mapper.replace_mapper_idx(conv_256_idx)
self.assertEqual(conv_256_idx.shape,
(chan_num, roi_property.BIOSEMI_CONV))