hzwer/ICCV2019-LearningToPaint
A model-based deep reinforcement learning system that teaches machines to paint images by learning stroke placement and color decisions.
Not currently ranked — collecting fresh signals.
star history
This research project uses deep RL with a neural renderer to train agents that decompose images into a small number of brush strokes. The agent learns to plan stroke positions and colors through trial and error without human painter demonstrations. It produces visual artwork by sequentially drawing strokes to reconstruct texture-rich images.
Frequently asked
- What is hzwer/ICCV2019-LearningToPaint?
- A model-based deep reinforcement learning system that teaches machines to paint images by learning stroke placement and color decisions.
- Is ICCV2019-LearningToPaint open source?
- Yes — hzwer/ICCV2019-LearningToPaint is open source, released under the MIT license.
- What language is ICCV2019-LearningToPaint written in?
- hzwer/ICCV2019-LearningToPaint is primarily written in Python.
- How popular is ICCV2019-LearningToPaint?
- hzwer/ICCV2019-LearningToPaint has 2.3k stars on GitHub.
- Where can I find ICCV2019-LearningToPaint?
- hzwer/ICCV2019-LearningToPaint is on GitHub at https://github.com/hzwer/ICCV2019-LearningToPaint.