Painter genetic algorithm.
- Python 3.8.6+
- Anaconda 4.9.2+ (optional)
- Python modules from the
requirements.txt
To run the genetic algorithm, use the command:
$ python .
The generated images can be fund in the paintings
directory.
To finetune the algorithm you can modify the parameters in the __main__.py
file.
This project is open for any kind of contribution from anyone.
- Fork this repository
- Create a new branch (optional)
- Clone it
- Make your changes
- Upload them
- Make a pull request here
These test were taken with these parameters:
POP_SIZE = 100 # Number of starting images
PERSON_SIZE = (32, 32,) # Size of images
The EPOCH
parameter varies row by row.
No. of evolution steps | Best (brightest) painting | Worst (darkest) painting | Time to finish |
---|---|---|---|
16 | 00:03 | ||
32 | 00:06 | ||
64 | 00:12 | ||
128 | 00:24 | ||
256 | 00:48 | ||
512 | 01:37 | ||
1024 | 03:18 |
On average 5.25 it/s.
Their improval with each step:
Copyright © 2022 Martin Kondor.
See the LICENSE file for details.