Connecting Images and Natural Language
☑ Attending!
Talk by Andrej Karpathy of Stanford
RSVP
Connecting Images and Natural LanguageTalk by Andrej Karpathy of Stanford
nlp (2 out of 2)
MSc by research Interdisciplinary Creative Practices
Edinburgh College of Art, University of Edinburgh
Core classes taught concepts from architecture, philosophy, art, sociology, media theory, political sciences, choreography, design, history, literature... and who knows what else. My eyes were opened to a million new ways of thinking about the world. I started to understand 'research by practice' and the value of play. There was still a lot of artsy crap I couldn't accept, but it was mostly okay. I took electives inside my comfort zone, from the School of Informatics (Advanced Natural Language Processing and Multi-agent Semantic Web Systems). I worked on group projects with an awesome and diverse bunch of people from whom I learnt a lot.
Thesis: Location-aware literature
(I cleverly wrote this thesis in Google Docs, but I might translate it to HTML and link to it one day).
I worked with a couple of people in the English Literature department to build Palimpsest, a web app (with Python and JavaScript on Google AppEngine) that shows you historical texts based on your location as you walk around a city. I spent most time on the authoring interface, for plotting texts onto a map. The project is still going, though it's morphed now and outgrown the code I wrote. I'd intended to develop this into a full blown interactive fiction engine and authoring system that let you draw the world map onto a real map, so that players could navigate the story by moving around IRL. I didn't pull that off, but I mocked it up.