Function deconv2D_naive

numpy_ml/neural_nets/utils/utils.py:720–794

Perform a "deconvolution" (more accurately, a transposed convolution) of an input volume `X` with a weight kernel `W`, incorporating stride, pad, and dilation. Notes ----- Rather than using the transpose of the convolution matrix, this approach uses a direct convolution

(X, W, stride, pad, dilation=0)

def deconv2D_naive(X, W, stride, pad, dilation=0):
    """
    Perform a "deconvolution" (more accurately, a transposed convolution) of an
    input volume `X` with a weight kernel `W`, incorporating stride, pad, and
    dilation.

    Notes
    -----
    Rather than using the transpose of the convolution matrix, this approach
    uses a direct convolution with zero padding, which, while conceptually
    straightforward, is computationally inefficient.

    For further explanation, see [1].

    References
    ----------
    .. [1] Dumoulin & Visin (2016). "A guide to convolution arithmetic for deep
       learning." https://arxiv.org/pdf/1603.07285v1.pdf

    Parameters
    ----------
    X : :py:class:`ndarray <numpy.ndarray>` of shape `(n_ex, in_rows, in_cols, in_ch)`
        Input volume (not padded)
    W: :py:class:`ndarray <numpy.ndarray>` of shape `(kernel_rows, kernel_cols, in_ch, out_ch)`
        A volume of convolution weights/kernels for a given layer
    stride : int
        The stride of each convolution kernel
    pad : tuple, int, or 'same'
        The padding amount. If 'same', add padding to ensure that the output of
        a 2D convolution with a kernel of `kernel_shape` and stride `stride`
        produces an output volume of the same dimensions as the input.  If
        2-tuple, specifies the number of padding rows and colums to add *on both
        sides* of the rows/columns in `X`. If 4-tuple, specifies the number of
        rows/columns to add to the top, bottom, left, and right of the input
        volume.
    dilation : int
        Number of pixels inserted between kernel elements. Default is 0.

    Returns
    -------
    Y : :py:class:`ndarray <numpy.ndarray>` of shape `(n_ex, out_rows, out_cols, n_out)`
        The decovolution of (padded) input volume `X` with `W` using stride `s` and
        dilation `d`.
    """
    if stride > 1:
        X = dilate(X, stride - 1)
        stride = 1

    # pad the input
    X_pad, p = pad2D(X, pad, W.shape[:2], stride=stride, dilation=dilation)

    n_ex, in_rows, in_cols, n_in = X_pad.shape
    fr, fc, n_in, n_out = W.shape
    s, d = stride, dilation
    pr1, pr2, pc1, pc2 = p

    # update effective filter shape based on dilation factor
    _fr, _fc = fr * (d + 1) - d, fc * (d + 1) - d