Hi Fred,
I installed the Theano version that contains  "Adding an AbstractConv3d 
interface #4862"  https://github.com/Theano/Theano/pull/4862 
but now  it doesn't work because in the previous Theano version, class Pool 
had these parameters: ds, ignore_border, st, padding, mode,openmp.
in the latest Theano version class Pool has no ds parameter: only  
ignore_border, mode,openmp.

In maxpool3d.py  I was calling op = DownsampleFactorMax((ds[1],ds[2]), 
ignore_border) where DownsampleFactorMax = pool.Pool

I tried   Pool(mode=..., ...)(input, ws=ds) but it doesn't work.
How can I call Pool passing (ds[1],ds[2]) ?

Many Thanks


You received this message because you are subscribed to the Google Groups 
"theano-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to theano-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
""" Max pooling spatio-temporal inputs for Theano """

from theano import tensor
from theano.tensor.signal.downsample import DownsampleFactorMax

#it was originally ignore_border=False and then corrected as suggested by Pascal
'''Pascal update on ignore_border'''   
def max_pool_3d(input, ds, ignore_border=True):
    Takes as input a N-D tensor, where N >= 3. It downscales the input video by
    the specified factor, by keeping only the maximum value of non-overlapping
    patches of size (ds[0],ds[1],ds[2]) (time, height, width)

    :type input: N-D theano tensor of input images.
    :param input: input images. Max pooling will be done over the 3 last dimensions.
    :type ds: tuple of length 3
    :param ds: factor by which to downscale. (2,2,2) will halve the video in each dimension.
    :param ignore_border: boolean value. When True, (5,5,5) input with ds=(2,2,2) will generate a
      (2,2,2) output. (3,3,3) otherwise.

    if input.ndim < 3:
        raise NotImplementedError('max_pool_3d requires a dimension >= 3')

    # extract nr dimensions
    vid_dim = input.ndim
    # max pool in two different steps, so we can use the 2d implementation of 
    # downsamplefactormax. First maxpool frames as usual. 
    # Then maxpool the time dimension. Shift the time dimension to the third 
    # position, so rows and cols are in the back

    # extract dimensions
    frame_shape = input.shape[-2:]
    # count the number of "leading" dimensions, store as dmatrix
    # tensor.prod: product of every term in x along axis
    batch_size = tensor.prod(input.shape[:-2])
    # Reshape x by right padding the shape with n_ones 1s. 
    batch_size = tensor.shape_padright(batch_size,1)
    # store as 4D tensor with shape: (batch_size,1,height,width)
    # Cast any tensor x to a Tensor of the same shape, but with a different numerical type dtype.
    new_shape = tensor.cast(tensor.join(0, batch_size,
                                        frame_shape), 'int32')
    input_4D = tensor.reshape(input, new_shape, ndim=4)

    # downsample mini-batch of videos in rows and cols
    op = DownsampleFactorMax((ds[1],ds[2]), ignore_border)
    output = op(input_4D)
    # restore to original shape
    outshape = tensor.join(0, input.shape[:-2], output.shape[-2:])
    out = tensor.reshape(output, outshape, ndim=input.ndim)

    # now maxpool time

    # output (time, rows, cols), reshape so that time is in the back
    shufl = (list(range(vid_dim-3)) + [vid_dim-2]+[vid_dim-1]+[vid_dim-3])
    input_time = out.dimshuffle(shufl)
    # reset dimensions
    vid_shape = input_time.shape[-2:]
    # count the number of "leading" dimensions, store as dmatrix    
    batch_size = tensor.prod(input_time.shape[:-2])
    batch_size = tensor.shape_padright(batch_size,1)
    # store as 4D tensor with shape: (batch_size,1,width,time)
    new_shape = tensor.cast(tensor.join(0, batch_size,
                                        vid_shape), 'int32')
    input_4D_time = tensor.reshape(input_time, new_shape, ndim=4)
    # downsample mini-batch of videos in time
    op = DownsampleFactorMax((1,ds[0]), ignore_border)
    outtime = op(input_4D_time)
    # output 
    # restore to original shape (xxx, rows, cols, time)
    outshape = tensor.join(0, input_time.shape[:-2], outtime.shape[-2:])
    shufl = (list(range(vid_dim-3)) + [vid_dim-1]+[vid_dim-3]+[vid_dim-2])
    return tensor.reshape(outtime, outshape, ndim=input.ndim).dimshuffle(shufl)

Reply via email to