Creative visual storytelling - that is, the creative task of storytelling based on visual input - involves both assigning meaning to the visual input and conveying that meaning in story form.  The resulting stories are more than literal descriptions of events or scenery: they contain narrative arcs with characters, goals, and conflicts in potentially endless circumstances. In this talk, I will lay out my research exploring the foundations of creative visual storytelling and automating this novel type of storytelling. I examine three properties critical to such systems and the narratives they generate: the systems are highly expressive, their productive capability is key to problem solving and establishing story frames; the systems are responsible, the narratives they generate are grounded in the source material and avoid biases; and the system narratives are "co-constructive" with a human partner, they enable interlocutors to share common ground of experiences in different physical spaces across time through evolving events.

Bio: Dr. Stephanie Lukin is a Computer Scientist at the U.S. Army Research Laboratory's Los Angeles regional site (ARL-West). She holds a Ph.D. and M.S. in Computer Science from the University of California Santa Cruz, and had interned at Xerox PARC, Microsoft Research, and Google before joining ARL in 2017. Dr. Lukin specializes in narrative intelligence, examining the multi-modal interactions between humans and robots, and how stories can be told from the myriad of multi-modal data surrounding us.