VIDEO NOW ONLINE: https://www.youtube.com/watch?v=2tUP_lTglpw
The field of chart question answering (CQA) has begun to combine analyses of images and text in visualizations using recent advances in computer vision and natural language processing.
Ideally, computational methods might mimic human chart perception, but how to best use computational approaches in ways that benefit from knowledge of human visual inference remains an open question, among many other unexplored research opportunities in CQA.
This workshop has three goals: (1) cognition researchers will introduce concepts of human chart perception to inspire algorithmic design, (2) computer scientists will present state-of-the-art machine learning techniques for chart analysis including respective limitations, (3) our proposed algorithmic challenge and call for submissions will spur the development of new CQA systems.
DATES
Recording HERE!
ORGANIZERS
 |
Daniel Haehn
University of Massachusetts Boston |
 |
Narges Mahyar
University of Massachusetts Amherst |
 |
Steven Franconeri
Northwestern University |
 |
Jessica Hullman
Northwestern University |
 |
Nikolaus Kriegeskorte
Columbia University |
 |
Hanspeter Pfister
Harvard University |
