Creates a video showing the AI learning a visual image. The video "output.mp4" is an example of a realtime video of an AI learning the example image "face.png".
Example Video (Not Real-time):
output_video.mp4
Reference image on left, final model ouput on right:
How To Use
- in AI_Learning.py, set the image file, and other option variables to what ever you want
- Then set the hyper parameters to what ever you want
- Then Run, and see the output video!
Notes:
- The realtime videos may have very High FPS, and some video players might slow them down to 60FPS which makes them non-realtime, so make sure to use a good video player
- When training the AI on different images you may need to tweak the Hyper parameters and the number of hidden layers
- Depending on the Size of the image and the hyper parameters you choose, you might want to run this on a GPU
- Finally, this is a project I made for fun. Obviously the Model could be improved massively using newer techniques like stable diffustion, and this is not how a typpical AI would be used. A normal AI that is developed is typically used to generalize information, not memorize like this model is. But again, thats the point, its just for fun :)