game generation

PokemonGym is a server-client architecture-based platform designed for AI agents to be evaluated and trained in the Pokemon Red game. It provides game state through FastAPI, supports human interaction with AI agents, and helps researchers and developers test and improve AI solutions.

SCENIC is a text-conditional scene interaction model that can adapt to complex scenes with different terrains and supports user-specified semantic control using natural language. The model navigates a 3D scene using user-specified trajectories as sub-goals and text prompts. SCENIC uses a hierarchical scene reasoning method, combined with frame alignment between motion and text, to achieve seamless transitions between different motion styles. The importance of this technology lies in its ability to generate character navigation actions that comply with real physical rules and user instructions. It is of great significance to fields such as virtual reality, augmented reality, and game development.

GameGen-X is a diffusion transformation model designed for generating and interactively controlling open-world game videos. This model achieves high-quality, open-domain video generation by simulating various features of game engines, such as innovative characters, dynamic environments, complex actions, and diverse events. In addition, it provides interactive control capabilities to predict and change future content based on current video clips, enabling gameplay simulation. To realize this vision, we first collected and built from scratch an Open World Video Game Dataset (OGameData), the first and largest open world game video generation and control dataset, containing more than 1 million diverse gameplay video clips from more than 150 games, all equipped with informative subtitles from GPT-4o. GameGen-X has gone through a two-stage training process, including basic model pre-training and instruction tuning. First, the model is pre-trained with text-to-video generation and video sequels, giving it the ability to generate long-sequence, high-quality open-domain game videos. Further, in order to achieve interactive control capabilities, we designed InstructNet to integrate game-related multi-modal control signal experts. This enables the model to adapt latent representations based on user input, unifying character interaction and scene content control in video generation for the first time. During instruction tuning, only InstructNet is updated, while the pre-trained base model is frozen, allowing the integration of interactive control capabilities without losing the diversity and quality of the generated video content. GameGen-X represents a major leap forward in using generative models for open-world video game design. It demonstrates the potential of generative models as an adjunct to traditional rendering techniques, effectively combining creative generation with interactive capabilities.

ALIEN is an artificial life simulation program based on CUDA's specialized physics and rendering engine. It is designed to simulate the behavior of digital organisms in artificial ecosystems and serve as a platform for evolutionary simulation. This software project is open source and follows the BSD-3-Clause license.

Tempest AI is a platform for quickly building games without coding. It uses a powerful AI assistant to help users create games from scratch, supporting dynamic characters, adaptive missions and interactive worlds. Users can build worlds through the drag-and-drop editor, design game mechanics, and control game narrative through the story builder. Tempest AI is not only suitable for game developers, but also for enthusiasts who want to explore creative and interactive storytelling.

AI Town is an MIT-licensed deployable starter kit for building and customizing your own version of AI Town. This project was inspired by the research paper 'Generative Agents: Interactive Simulacra of Human Behavior' and aims to provide a powerful base platform designed to be extended. The backend natively supports shared global state, transactions, and a simulation engine for everything from simple projects to scalable multiplayer games.

Buenoverse is a no-code 3D world and game building platform for ordinary users. It provides a drag-and-drop visual editor and integrates a large number of 3D resources. Users can quickly build their own 3D spaces, scenes and even games without programming. The platform supports browsers and VR devices, and the output works can be used on websites or apps, or can be packaged to generate executable programs.

Betafish.js is a JavaScript library that provides chess AI functionality. It can perform game analysis based on the incoming FEN string and generate the next best AI move. Betafish.js has an efficient search algorithm and optimized evaluation function to make the best decision within a given thinking time. It is suitable for building online chess games, educational projects or for testing and demonstrating AI algorithms.