vizdoom 1.3.2

ViZDoom allows developing AI bots that play Doom using only visual information (the screen buffer). It is primarily intended for research in machine visual learning, and deep reinforcement learning, in particular.

ViZDoom is based on ZDoom engine to provide the game mechanics.

Features

• API for Python (including Gymnasium/Gym wrappers) and C++,
• Multi-platform (Linux, macOS, Windows),
• Fast (up to 7000 frames/steps per second in sync mode, single-threaded on a modern CPU),
• Lightweight (few MBs),
• Easy-to-create custom scenarios (visual editors, powerful scripting language, and examples available),
• Async and sync single-player and multiplayer modes,
• Customizable resolution and rendering parameters,
• Access to the depth buffer (3D vision),
• Automatic labeling and categorization of game objects visible in the frame,
• Access to the audio buffer,
• Access to the list of actors/objects and map geometry,
• Access to in-game text messages and notifications,
• Off-screen rendering,
• Episodes recording,
• In-game time scaling in async mode.

ViZDoom API is reinforcement learning friendly (suitable also for learning from demonstration, apprenticeship learning or apprenticeship via inverse reinforcement learning, etc.).

Cite as

M Wydmuch, M Kempka & W Jaśkowski, ViZDoom Competitions: Playing Doom from Pixels, IEEE Transactions on Games, vol. 11, no. 3, pp. 248-259, 2019 (arXiv:1809.03470)

@article{Wydmuch2019ViZdoom,
  author  = {Marek Wydmuch and Micha{\l} Kempka and Wojciech Ja\'skowski},
  title   = {{ViZDoom} {C}ompetitions: {P}laying {D}oom from {P}ixels},
  journal = {IEEE Transactions on Games},
  year    = {2019},
  volume  = {11},
  number  = {3},
  pages   = {248--259},
  doi     = {10.1109/TG.2018.2877047},
  note    = {The 2022 IEEE Transactions on Games Outstanding Paper Award}
}

or/and

M. Kempka, M. Wydmuch, G. Runc, J. Toczek & W. Jaśkowski, ViZDoom: A Doom-based AI Research Platform for Visual Reinforcement Learning, IEEE Conference on Computational Intelligence and Games, pp. 341-348, Santorini, Greece, 2016 (arXiv:1605.02097)

@inproceedings{Kempka2016ViZDoom,
  author    = {Micha{\l} Kempka and Marek Wydmuch and Grzegorz Runc and Jakub Toczek and Wojciech Ja\'skowski},
  title     = {{ViZDoom}: A {D}oom-based {AI} Research Platform for Visual Reinforcement Learning},
  booktitle = {IEEE Conference on Computational Intelligence and Games},
  year      = {2016},
  address   = {Santorini, Greece},
  month     = {Sep},
  pages     = {341--348},
  publisher = {IEEE},
  doi       = {10.1109/CIG.2016.7860433},
  note      = {The Best Paper Award}
}

Python quick start

Linux

To install the latest release of ViZDoom, just run:

pip install vizdoom

Both x86-64 and AArch64 (ARM64) architectures are supported. Wheels are available for Python 3.10+ on Linux.

⚠️ To use audio features, you need OpenAL install in your system. On apt-based distros (Ubuntu, Debian, Linux Mint, etc.)

apt install libopenal-dev

On dnf/yum-based distros (Fedora, RHEL, CentOS, Alma/Rocky Linux, etc.)

dnf install openal-soft-devel

If Python wheel is not available for your platform (Python version <3.10, distros below manylinux_2_28 standard), pip will try to install (build) ViZDoom from the source. ViZDoom requires a C++11 compiler, CMake 3.12+, Boost 1.54+ SDL2, OpenAL (optional) to install from source. See documentation for more details.

macOS

To install the latest release of ViZDoom, just run:

pip install vizdoom

Since 1.3.0+, pre-build wheels are available only for Apple Silicon (M-series chips) macOS 14.0+.

⚠️ To install pre-build wheels on Intel macOS 13.0+, you need to install version 1.2.4 using pip:

pip install vizdoom==1.2.4

If Python wheel is not available for your platform (Python version <3.10, older macOS version), pip will try to install (build) ViZDoom from the source. ViZDoom requires a C++11 compiler, CMake 3.12+, Boost 1.54+ SDL2, OpenAL (optional) to install from source. See documentation for more details how to install dependencies.

Windows

To install the latest release of ViZDoom, just run:

pip install vizdoom

At the moment, only x86-64 architecture is supported on Windows. Wheels are available for Python 3.10+ on Windows.

Please note that the Windows version is not as well-tested as Linux and macOS versions. It can be used for development and testing but if you want to conduct serious (time and resource-extensive) experiments on Windows, please consider using Docker or WSL with Linux version.

Gymnasium/Gym wrappers

Gymnasium environments are installed along with ViZDoom and are available on all platforms. See documentation and examples on the use of Gymnasium API.

Examples

• Python (contain learning examples implemented in PyTorch, TensorFlow, and Theano)
• C++

Python examples are currently the richest, so we recommend looking at them, even if you plan to use C++. The API is almost identical between the languages, with the only difference being that Python uses snake_case and C++ camelCase for methods and functions.

Original Doom graphics

Unfortunately, we cannot distribute ViZDoom with original Doom graphics. If you own original Doom and Doom 2 games, you can replace Freedoom graphics by placing doom2.wad into your working directory or vizdoom package directory.

Alternatively, any base game WAD (including other Doom engine-based games and custom/community games) can be used by pointing to it with the set_doom_game_path/setDoomGamePath method.

Documentation

Detailed descriptions of all ViZDoom types and methods can be found in the documentation.

Full documentation of the ZDoom engine and ACS scripting language can be found on ZDoom Wiki.

Useful articles (for advanced users who want to create custom environments/scenarios):

• ZDoom Wiki: ACS (scripting language)
• ZDoom Wiki: CVARs (console variables)
• ZDoom Wiki: CCMD (console commands)

Awesome Doom tools/projects

• SLADE3 - Great Doom map (scenario) editor for Linux, macOS and Windows.
• Doom Builder 2 - Another great Doom map editor for Windows.
• OBLIGE - Doom random map generator and PyOblige is a simple Python wrapper for it.
• Omgifol - Nice Python library for manipulating Doom maps.
• NavDoom - Maze navigation generator for ViZDoom (similar to DeepMind Lab).
• MazeExplorer - A more sophisticated maze navigation generator for ViZDoom.
• Sample Factory - A high-performance reinforcement learning framework for ViZDoom.
• EnvPool - A high-performance vectorized environment for ViZDoom.
• Obsidian - Doom random map generator, a continuation of OBLIGE.
• LevDoom - Generalization benchmark in ViZDoom featuring difficulty levels in visual complexity.
• COOM - Continual learning benchmark in ViZDoom offering task sequences with diverse objectives.
• HASARD - A safe reinforcement learning benchmark in ViZDoom

If you have a cool project that uses ViZDoom or could be interesting to ViZDoom community, feel free to open PR to add it to this list!

Contributions

This project is maintained and developed in our free time. All bug fixes, new examples, scenarios, and other contributions are welcome! We are also open to feature ideas and design suggestions.

We have a roadmap for future development work for ViZDoom available here.

License

The code original to ViZDoom is under MIT license. ZDoom uses code from several sources with varying licensing schemes.