Learning to Paint With Model-Based Deep Reinforcement Learning

Citations Over TimeTop 10% of 2019 papers

Abstract

We show how to teach machines to paint like human painters, who can use a small number of strokes to create fantastic paintings. By employing a neural renderer in model-based Deep Reinforcement Learning (DRL), our agents learn to determine the position and color of each stroke and make long-term plans to decompose texture-rich images into strokes. Experiments demonstrate that excellent visual effects can be achieved using hundreds of strokes. The training process does not require the experience of human painters or stroke tracking data. The code is available at https://github.com/hzwer/ICCV2019-LearningToPaint.

Related Papers

• → Re‐Evaluating Court and Folk Painting of Korea(2011)2 cited
• Elementary Disussion on Painting Comments and History(2002)
• On the Artistic Features of Intellectual's Paintings in Song Dynaty(2009)
• The coarse painter and his position in 17th- and 18th-century Dutch decorative painting(2015)
• → The Classification System of Painting Divisions and Painting Subjects(2018)