»Towards content aware image search in GNU/Linux«
2018-10-14, 12:40–13:10, Α1
A talk focused on Computer Vision, its applications in Linux and how we can harness the power of deep convolutional neural networks to perform content aware image search on our Linux desktops and servers.
This talk will give you some insight on the current computer vision revolution happening with deep neural networks and how they can be harnessed to power our linux desktops and servers.
We will examine a brief history of the effort of solving the computer vision problem while not delving in too many technical details, mention free and opensource toolkits widely used, their underlying similarities and differences and the current state of the art.
Finally we will propose an open-source toolkit that is currently just a draft prototype and proof of concept but that can eventually allow UNIX systems to yield the same level of power we are now starting to see in services like google image/ search etc from the comfort of our command line interface.