W zależności od ilości danych do przetworzenia generowanie pliku może się wydłużyć.

Jeśli generowanie trwa zbyt długo można ograniczyć dane np. zmniejszając zakres lat.

Rozdział

Pobierz BibTeX

Tytuł

ViZDoom: A Doom-based AI Research Platform for Visual Reinforcement Learning

Autorzy

[ 1 ] Instytut Informatyki, Wydział Informatyki, Politechnika Poznańska | [ P ] pracownik

Rok publikacji

2016

Typ rozdziału

referat

Język publikacji

angielski

Słowa kluczowe
EN
  • video games
  • visual-based reinforcement learning
  • deep reinforcement learning
  • first-person perspective games
  • FPS
  • visual learning
  • neural networks
Streszczenie

EN The recent advances in deep neural networks have led to effective vision-based reinforcement learning methods that have been employed to obtain human-level controllers in Atari 2600 games from pixel data. Atari 2600 games, however, do not resemble real-world tasks since they involve non-realistic 2D environments and the third-person perspective. Here, we propose a novel test-bed platform for reinforcement learning research from raw visual information which employs the first-person perspective in a semi-realistic 3D world. The software, called ViZDoom, is based on the classical first-person shooter video game, Doom. It allows developing bots that play the game using the screen buffer. ViZDoom is lightweight, fast, and highly customizable via a convenient mechanism of user scenarios. In the experimental part, we test the environment by trying to learn bots for two scenarios: a basic move-and-shoot task and a more complex maze-navigation problem. Using convolutional deep neural networks with Q-learning and experience replay, for both scenarios, we were able to train competent bots, which exhibit human-like behaviors. The results confirm the utility of ViZDoom as an AI research platform and imply that visual reinforcement learning in 3D realistic first-person perspective environments is feasible.

DOI

10.1109/CIG.2016.7860433

URL

https://ieeexplore.ieee.org/document/7860433

Książka

Proceedings of IEEE Conference on Computational Intelligence and Games (CIG 2016)

Zaprezentowany na

IEEE Conference on Computational Intelligence and Games, CIG 2016, 20-23.09.2016, Santorini, Greece

Publikacja indeksowana w

WoS (15)

Ta strona używa plików Cookies, w celu zapamiętania uwierzytelnionej sesji użytkownika. Aby dowiedzieć się więcej przeczytaj o plikach Cookies i Polityce Prywatności.