Stack Overflow
Stack Overflow

Questions tagged [deconvolution]

An algorithmic process to reverse the effects of a convolution, which is a linear form of signal or image filtering.

Deconvolution is the process of estimating an original signal (or image) from recorded data. Usually, the process of recording the signal involves a transfer function (in imaging, a point spread function), which is convolved with the original signal and blurs it. Deconvolution can then be applied to improve the signal quality. For example, in fluorescence microscopy deconvolution can be used together with special illumination configurations to obtain super-resolved images. If the transfer function cannot be measured, bind deconvolution can be used to estimate both the original signal and the transfer function simultaneously.

Deconvolution typically involves regularization and iterative optimization methods.

Related tags

203 questions
7
votes
2 answers

Tensorflow equivalent of the Keras function: UpSampling2D

I would like to use the Keras layer: from keras.layers.convolutional import UpSampling2D x = UpSampling2D((2, 2))(x) How can I replicate this behavior with native tensorflow ? I can't manage to find an equivalent function/layer.
Jonathan DEKHTIAR
  • 3,456
  • 1
  • 21
  • 42
6
votes
1 answer

In Keras what is the difference between Conv2DTranspose and Conv2D

I'm currently building a GAN with Tensorflow 2 and Keras and noticed a lot of the existing Neural Networks for the generator and discriminator use Conv2D and Conv2DTranspose in Keras. I'm struggling to find something that functionally explains the…
Zapster
  • 71
  • 1
  • 7
6
votes
1 answer

MetalKit for iOS 10 : Execution of the command buffer was aborted due to an error during execution. Caused GPU Timeout Error (IOAF Code 2)

Using MetalKit for iOS 10, when we try to perform MPSCNNConvolution, with inputs as following : Kernel Size : 16x16 Input channels : 300 Output channels : 250 Dimensions of input image : 250x250x300 Execution of Command Buffer takes over 10 seconds…
mayuur
  • 4,736
  • 4
  • 30
  • 65
6
votes
1 answer

How to stack multiple layers of conv2d_transpose() of Tensorflow

I'm trying to stack 2 layers of tf.nn.conv2d_transpose() to up-sample a tensor. It works fine during feed forward, but I get an error during backward propagation: ValueError: Incompatible shapes for broadcasting: (8, 256, 256, 24) and (8, 100, 100,…
Pette
  • 155
  • 2
  • 8
5
votes
2 answers

Use Python lmfit with a variable number of parameters in function

I am trying to deconvolve complex gas chromatogram signals into individual gaussian signals. Here is an example, where the dotted line represents the signal I am trying to deconvolve. I was able to write the code to do this using…
user14241
  • 727
  • 1
  • 8
  • 27
5
votes
0 answers

My Image segmentation result map contains black lattice in in the white patch

I'm doing an image segmentation with UNet-like CNN architecture by Pytorch 0.4.0.It mark foreground as 1 and background as 0 in the final segmentation result.I use a pre-trained VGG's feature extractor as my encoder, so I need to upsampling the…
Kevin
  • 71
  • 5
5
votes
1 answer

What does the padding of tf.nn.conv2d_transpose do?

As we know, we can calculate the shape of output tensor by padding mode for conv2d, and the algorithm is clear, but I'm very confused about conv2d_transpose, does it pad the input tensor and then invoke conv2d? And where does it transpose filter or…
gaussclb
  • 1,217
  • 3
  • 13
  • 26
5
votes
1 answer

How to increase the size of deconv2d filters for a fixed data size?

I am trying to adjust this DCGAN code to be able to work with 2x80 data samples. All generator layers are tf.nn.deconv2d other than h0, which is ReLu. Generator filter sizes per level are currently: Generator: h0: s_h16 x s_w16: 1 x 5 Generator:…
Massyanya
  • 2,844
  • 8
  • 28
  • 37
5
votes
1 answer

deblurring image by deconvolution using opencv

I have two images o1 & o2, and I have blurred the two images using the same Gaussian blurring kernel. Then I have found kernel k1 = DFT(b1) / DFT (o1), where b1 is the image obtained by blurring o1. I have used this kernal (k1) to perform…
Deepak
  • 1,038
  • 5
  • 17
  • 44
4
votes
1 answer

sqrt but for convolution

I've got a distribution g that is the result of a signal convolved with itself f * f = g. I'd like to extract f using a deconvolution, but scipy.signal.deconvolve requires knowledge of the filter, which in this case is simply f. How might I go about…
Ned Bingham
  • 2,749
  • 18
  • 23
4
votes
3 answers

Tensorflow conv2d_transpose: Size of out_backprop doesn't match computed

When I build the FCN for segmentation, I want the images to keep the original size of input data, so I use the fully convolution layers. When I choose the fixed input size, such as (224, 224), the transpose conv works fine. However, when I changed…
Cwang
  • 371
  • 3
  • 9
4
votes
1 answer

Visualization of the filters of VGG16

I am learning CNN, right now, working on deconvolution of the layers. I have begun the process of learning upsampling and observe how convolution layers see the world by generating feature maps from the filters from the source Visualization of the…
sana
  • 410
  • 2
  • 6
  • 24
4
votes
0 answers

Practical difference between deconvolution and resize followed by convolution for upscaling

I need to go (upsample) from a layer input = HxWxn1, where H: height, W:width and n1: number of filters, to a layer output = 2Hx2Wxn2, where 2H = 2*Height etc, and n2=n1/2: number of new filters. One way of achieving this is by using transposed…
Foivos
  • 545
  • 4
  • 13
4
votes
0 answers

Deconvolution does not work on real data (even if they are up to few decimals similar to modeled)

I have encountered problem with pseudo-real data deconvolution. What I want to do is to create some pulse profile, convolve with gaussian, add some little noise and deconvolve. However this adding of slight discrepancy (having finite number of…
Petr Synek
  • 143
  • 1
  • 8
4
votes
1 answer

How to visualize intermediate feature layers in keras?

I've read the paper Visualizing and Understanding Convolutional Networks by Zeiler and Fergus and would like to make use of their visualization technique. The paper sounds promising - but unfortunately, I have no idea how to implement it in Keras…
D.Laupheimer
  • 1,074
  • 1
  • 9
  • 21
1
2
3
…
13 14