-
Extend libBrain library with all Convolutin Network object based on this
http://neuralnetworksanddeeplearning.com/chap6.html
http://ufldl.stanford.edu/tutorial/supervised/ConvolutionalNeuralNetwork/…
-
For the 2nd milestone, we would like to make a start at using real galaxies for simulations. To do this, we have to decide how those data will be stored. Examples of decisions to be made:
- do we st…
-
Right now we are limited to Sersic n in the range [0.5, 4.2]. What would be required to go to higher n? A non-negligible fraction of galaxies in HST imaging are fit with higher n, up to ~7 or so. M…
-
Note if our convolutions are occurring in spatial domain we are likely to be fairly efficient at present.
Opportunity for speed up in ConvolveFilterBankTo2D as linear sum. This would save a few funct…
-
Hi,
The most difficult function could be the FFT:
- http://docs.nvidia.com/cuda/cufft/#multi-dimensional
- FFT can segfault a lot due to constraints such as dimension along you want to apply it and m…
-
Hi guys,
I just updated my Gadgetron to the latest master branch, and I noticed this commit "dde8b6d56002c376f29becc982c8bc0e221d11ca" with commit message "Updated cuNDFFT to use correct scaling and u…
-
-
I found a minor issue with the number of steps required to compute the best period. It seems that the number of steps scales linearly with the scale of the time units. A simple workaround is to just u…
-
Hello again,
I'm having a problem since the last few days. Let's say I have a really simple code:
``` python
import numpy as np
np.
```
After some time processing for the autocompletion the next me…
-
_Motivation:_ We are interested in speeding up the rendering of galaxy images for use in MCMC algorithms to infer galaxy model parameters given a postage stamp image. An alternative galaxy model param…