/Paper-2020-SBGames

Logs and trained models of the paper "Investigating Deep Q-Network Agent Sensibility to Texture Changes on FPS Games" presented at SBGames 2020.

Investigating Deep Q-Network Agent Sensibility to Texture Changes on FPS Games

Conference: Brazilian Symposium on Computer Games and Digital Entertainment (SBGames)

Workflow

[Page] [PDF]

Folders

Results

Results

Fixed Maps: The monster has predefined spawn points shared across all scenarios.
Basic: At every new episode the monster spawns in a random place.

Scenarios

Animated: custom scenario with animated walls, ceiling, and floor.
Basic: standard Basic scenario from VizDoom.
Caco: custom scenario with the original Cacodemon monster skin.
Flat: custom scenario with only flat textures.

Training Logs

Full training logs of the agents in all experiments.

Trained Models

You can find the four trained agents in this folder.
To load a model, extract the files then use the Keras command load_model.

Cite this paper

BibTex

@InProceedings{serafim2020investigating,
  title = {Investigating Deep Q-Network Agent Sensibility to Texture Changes on {FPS} Games},
  author  = {Serafim, Paulo Bruno Sousa and Nogueira, Yuri Lenon Barbosa and Vidal, Creto Augusto and Cavalcante-Neto, Joaquim Bento and F\'{e}rrer Filho, R\^{o}mulo Freire},
  booktitle = {Proceedings of the XIX Brazilian Symposium on Computer Games and Digital Entertainment (SBGames)},
  pages = {117--125},
  year = {2020},
  issn = {2179-2259},
  doi = {10.1109/SBGames51465.2020.00025}
}