Skip to content

Latest commit

 

History

1 Commit

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Generating Heider-Simmel animations used for experiments

Requirements

  • Python 3.5
  • Numpy >= 1.14.2
  • PyTorch 0.3.1
  • pybox2d
  • pygame
  • OpenCV >= 3.4.2

HH

Style 1, 100% AD

python -m experiments.gen_social --exp-name Blocking_v1 --checkpoints 2000 16000 --action-freq 1

Style 2, 100% AD

python -m experiments.gen_social --exp-name Blocking_v2 --checkpoints 34400 16000 --action-freq 1

Change AD

Change the argument action-freq from 1, 2, 5, 10, to 15 for 100% AD, 50% AD, 20% AD, 10% AD, and 7% AD respectively

HO

Style 1

python -m experiments.gen_social --exp-name Blocking_v0 --checkpoints 0 16000

Style 2

python -m experiments.gen_social --exp-name Blocking_v0 --checkpoints 0 46000 

Style 3

python -m experiments.gen_social --exp-name Blocking_v0 --checkpoints 0 16000 --restitution 1

OO

collision

python -m experiments.gen_physical --exp-name collision

rod

python -m experiments.gen_physical --exp-name rod

rope

python -m experiments.gen_physical --exp-name rope

spring

python -m experiments.gen_physical --exp-name rope

About

Generating Heider-Simmel type of annimations using reinforcement learning

Resources

Stars

5 stars

Watchers

1 watching

Forks

Releases

Packages

Contributors

Languages