-
Hi guys,
Are convolution algorithm or fast Fourier transform implemented to perform filtering or frequency analysis?
It could be of great interest.
Kind Regards,
Luis
-
Add algorithms useful for vision / signal processing:
- [ ] Convolution (with implicit im2col)
- [ ] Conv transpose
- [ ] Fast Fourier Transform
-
Implement Fast Fourier Transformations for convolutions
-
## 🚀 Feature
An FFT backend for `kornia.filters.filter2d`.
## Motivation
`kornia.filters.filter2d` is very slow for large kernels as it currently only performs convolutions in the spatial domai…
-
So, I had this idea that if we can modify pixel manipulation api to support separable filters, then we can speed up our performance for stuff like edge detection quite a bit.
Since that takes us from…
-
Hello there, I want to clarify whether the need for square matrices is strictly enforced. From the paper, I note that
"We turn to an expressive class of sub-quadratic structured matrices called Mo…
-
clever solution that immensely speeds up the expected_cis calculation is to do fast Fourier convolution for calculation of pairwise distances between bad bins ([here](https://github.com/open2c/cooltoo…
-
Hi all,
It's great to see that spectral cube now has a `convolve_to` method - this is going to be super useful. I see, though, that by default `astropy` convolution is used. I've encountered issues…
-
I have written a hacky script to look for dead links. Open a console and `cd path/to/fundamentals_of_interferometry`. Then fire up ipython and copy the below code (ctrl-c):
```python
import json
…
-
Is there an equivalent for the Python FFt function fftconvolve(in1, in2) from the scipy library?
I need this for 2 equal sized 2D arrays.
Unfortunately I couldn't find anything suitable.
Here is …