Computer vision is making rapid progress in its ability to extract information from documents. Libraries, academic digital humanists and publishers are now moving from experimenting with computer vision to considering how to implement it in their search platforms or research methods, or support it for their users.
This one-day symposium will combine presentations on current research using computer vision with hands-on training in computer vision tools. It will both showcase the state-of-the-art in the field and discuss how it can be made more accessible to users of libraries and digital repositories.
For the full workshop programme, please see the event's website at https://www.cerl.org/services/seminars/cvinbookhistory
You can register for the whole day, only the morning workshop (maximum number of participants is 25), or the afternoon programme with lightning talks, key note presentation and discussion (max. 50).
If you wish to attend the morning workshop we ask you to explain in one or two sentences why it would be important for you and your research to attend. If we receive more than 25 registrations, we will select participants on the basis of the reasons you gave.
No prior knowledge of computer vision is required. Participants should bring their own laptop, ideally one on which they have installed Docker in advance (see instructions for Mac, Windows or Linux in the section ‘Docker Version’ at http://www.robots.ox.ac.uk/~vgg/software/vic/). The workshop can, however, be followed without installing Docker.
This event is co-organised with CERL, Brill and the Visual Geometry Group of the University of Oxford.