Title:	Generative models for visual storytelling
DNr:	Berzelius-2024-396
Project Type:	LiU Berzelius
Principal Investigator:	Daniel Jönsson <daniel.jonsson@liu.se>
Affiliation:	Linköpings universitet
Duration:	2024-10-06 – 2025-05-01
Classification:	10209
Keywords:

Abstract

The goal of the project is to create visual summaries of stories using large lange and text-to-image models. These summaries are intended to consist of visually rich media, such as a series of images or videos complemented by text. Such summaries will enable alternative ways of experiencing written media. Efficient creation of the visual summaries will be enabled by the use of generate models.