ascend-tools/pt2pb/onnx-tensorflow/onnx_tf/handlers/backend/unpool_mixin.py

import tensorflow as tf

from onnx_tf.common import get_data_format
from onnx_tf.common import get_perm_from_formats


class UnpoolMixin(object):

    @classmethod
    def max_unpool(cls, node, input_dict):
        """
            MaxUnpooling operation
        """
        x = input_dict[node.inputs[0]]
        ind = input_dict[node.inputs[1]]
        if len(node.inputs) > 2:
            output_shape = input_dict.get(node.inputs[2], None)
        else:
            output_shape = None

        input_shape = x.get_shape()
        x_rank = len(x.get_shape())
        spatial_size = x_rank - 2
        storage_format, _ = get_data_format(x_rank)

        kernel_shape = node.attrs["kernel_shape"]
        # if strides are not provided default is 1 along each spatial axis
        strides = node.attrs.get("strides", [1] * spatial_size)
        pads = node.attrs.get("pads", None)

        default_shape = cls._get_default_shape(input_shape, kernel_shape,
                                               strides)

        need_trans = storage_format != "NHWC"
        if need_trans:
            x = tf.transpose(x, perm=get_perm_from_formats(storage_format,
                             "NHWC"))
            ind = tf.transpose(ind, perm=get_perm_from_formats(storage_format,
                               "NHWC"))

        # default_shape to NHWC storage format
        default_shape = [int(input_shape[0])] + default_shape + \
                        [int(input_shape[1])]

        unpooled = cls._unpool(x, ind, default_shape)

        if need_trans:
            unpooled = tf.transpose(
                unpooled, perm=get_perm_from_formats("NHWC", storage_format))

        if output_shape is not None:
            pads = cls._get_pads_from_output_shape(unpooled, output_shape)
        if pads is not None:
            unpooled = cls._pad_output(unpooled, pads, 0)

        return [unpooled]

    @classmethod
    def _get_default_shape(cls, input_shape, kernel_shape, strides):
        """
            Calculates default shape from kernel_shape and strides
            Args:
                input_shape:   shape of the input to unpool op
                kernel_shape:  the size of the kernel along each axis
                output_shape:  stride along each spatial axis
          Return:
            default_shape: calculated default_shape
        """
        default_shape = []
        for d in range(len(kernel_shape)):
            default_shape.append((int(input_shape[d + 2]) - 1) *
                                 int(strides[d]) + int(kernel_shape[d]))
        return default_shape

    @classmethod
    def _get_pads_from_output_shape(cls, unpool, output_shape):
        """
            Calculates the paddings from specified output_shape
            Args:
                unpool:       result from unpool operation
                output_shape: expected shape of the output
            Return:
                pads:         calculated paddings in format
                              [x1_begin, x2_begin,.., x1_end, x2_end]
                              where xi_... represent pads added to begin
                              or end of axis i
        """
        unpool_shape = tf.cast(tf.shape(unpool), dtype=tf.int32)
        new_shape = tf.cast(output_shape, dtype=tf.int32)

        pads_begin = []
        pads_end = []

        for d in range(len(unpool.get_shape())):
            pad_total = new_shape[d] - unpool_shape[d]
            pad_begin = tf.cast(pad_total / 2, tf.int32)
            pad_end = pad_total - pad_begin
            pads_begin = pads_begin + [pad_begin]
            pads_end = pads_end + [pad_end]

        pads = pads_begin + pads_end
        return pads

    @classmethod
    def _pad_output(cls, unpool, pads, constant_values):
        """
            Pad the output from unpool op
            Args:
                unpool:         result from unpool op
                pads:           paddings in format
                                [x1_begin, x2_begin,..., x1_end, x2_end]
                constant_values: constant value to fill up the padded spaces
            Return:
                padded:         padded tensor
        """
        unpool_shape = unpool.get_shape()
        paddings = []
        for d in range(len(unpool_shape)):
            paddings = paddings + [[pads[d], pads[d + len(unpool_shape)]]]
        padded = tf.pad(unpool, paddings, 'CONSTANT',
                        constant_values=constant_values)
        return padded

    @classmethod
    def _unpool(cls, pool, ind, output_shape, scope='unpool'):
        """
            Unpooling layer after max_pool_with_argmax.

            Args:
                pool:          max pooled output tensor
                ind:           argmax indices
                output_shape:  the shape of the output
            Return:
                unpool:        unpooling tensor
        """
        with tf.variable_scope(scope):
            input_shape = tf.shape(pool)

            flat_input_size = tf.reduce_prod(input_shape)
            flat_output_shape = [output_shape[0], output_shape[1] *
                                 output_shape[2] * output_shape[3]]

            pool_ = tf.reshape(pool, [flat_input_size])
            batch_range = tf.reshape(
                    tf.range(tf.cast(output_shape[0], tf.int64),
                             dtype=ind.dtype), shape=[input_shape[0], 1, 1, 1])
            b = tf.ones_like(ind) * batch_range
            b1 = tf.reshape(b, [flat_input_size, 1])
            ind_ = tf.reshape(ind, [flat_input_size, 1])
            ind_ = tf.concat([b1, ind_], 1)

            ret = tf.scatter_nd(ind_, pool_, shape=tf.cast(flat_output_shape,
                                                           tf.int64))
            ret = tf.reshape(ret, output_shape)
        return ret