Learning Multimodal Affinities for Textual Editing in Images
Event Type
Technical Paper
This session WILL be recorded.
Interest Areas
Research & Education
Registration Levels
Ultimate Supporter
Ultimate Attendee
Exhibitor Ultimate
TimeWednesday, 11 August 20217am - 8am PDT
DescriptionWe present an unsupervised method to learn multimodal affinities of textual entities in document images, considering their style, syntax, semantics, and geometry, and we demonstrate its applicability to various editing operations.