Page 1
Standard

Neural Artistic Style using Torch in Mint

Last year, this paper was published on "A Neural Algorithm of Artistic Style", which addresses how a neural net can replicate an artist's style and apply it to a photograph. All the pictures that float about on the web look amazing and I wanted to give this method a go, too. Unfortunately, as this is a pretty recent publication, there is no off-the-shelf program that implements this tool, and the Git repos I found were not very clear to set up in a VM:

Now, after a bit of tinkering around I managed to get the first of the repos working, although I would have loved to run the Python project.

Continue Reading →

The end
Standard

Nvidia on Ubuntu 14.04 LTS

I already wrote a quick article about how I managed to install the Nvidia drivers on Ubuntu 13.10, but with the latest LTS version it turned out to be a slightly different approach. One could say easier, but maybe that's due to me having done it so often and hence having gotten used to the procedure. In Ubuntu 13.10 I also discovered, that upon reboot the drivers intermittently decided to no longer work. The only fix I found so far was a complete purge-remove followed by a re-installation of said drivers. Now I believe to have found a fix  Continue Reading →

The end
Standard

Getting OpenCL (1.1) running in Ubuntu 13.10+

OpenCL is a great language for writing code that execute in parallel on a heterogeneous computing system. Such systems refer to electronic systems that use a variety of different types of computational units. These units include your central processing unit (CPU) and any available platform capable of executing commands that is part of the system. Over the years, OpenCL has become a "sort of industrial" standard and is supported by several hardware manufactures. Since I am still eager to use the Parallella boards and also am an enthusiast for computation and software implementation/simulation of algorithms, I decided to learn OpenCL. My current PC build therefore has two nice graphics cards (that I could afford) and I am hoping to use both in parallel with OpenCL. My developing OS of choice is Ubuntu since its graphical user interface (GUI) can be switched off, freeing a lot of memory; plus Ubuntu is a free Linux distribution and many specialized development systems like the Raspberry Pi or Parallella run on a Linux distro. So in case there is somebody else out there who may want to do some OpenCL development on Ubuntu, here comes a nice explanation how to set up everything.

Continue Reading →

The end
Standard

Nvidia on Ubuntu 13.10+

First I should apologise for the slow updates since I promised to have my tutorials done by now. I must confess that the C tutorial has been sitting in my unpublished post stack for way too long. If you are intrigued what I have done over the past months; I moved to Edinburgh, started my doctorate, learnt a whole module of mechanics & hydrodynamics and now I am also learning OpenCL (the latter subject for my own interest). Yet apologies are in order, I could have crammed in an hour or two of writing.

Continue Reading →

The end