1 paper · 1 filter
Jonas Belouadi, Eddy Ilg, Margret Keuper +5
Automatically synthesizing figures from text captions is a compelling capability. However, achieving high geometric precision and editability requires representing figures as graph…