Summary of Gamegen-x: Interactive Open-world Game Video Generation, by Haoxuan Che et al.
GameGen-X: Interactive Open-world Game Video Generation
by Haoxuan Che, Xuanhua He, Quande Liu, Cheng Jin, Hao Chen
First submitted to arxiv on: 1 Nov 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed GameGen-X model is a diffusion transformer designed for generating and interactively controlling open-world game videos. It simulates various game engine features, such as characters, environments, actions, and events, allowing for high-quality open-domain generation. The model also provides interactive controllability, predicting and altering future content based on the current clip. To achieve this, GameGen-X undergoes a two-stage training process, comprising foundation model pre-training and instruction tuning. The model is trained on an Open-World Video Game Dataset, the largest dataset for open-world game video generation and control, which consists of over a million diverse gameplay video clips from over 150 games with informative captions. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The GameGen-X model can generate high-quality open-domain game videos by simulating various game engine features. It also allows users to interactively control the generated content. This is achieved through a two-stage training process, which includes pre-training and instruction tuning. The model uses an Open-World Video Game Dataset, which is the largest dataset for this type of task. |
Keywords
» Artificial intelligence » Diffusion » Instruction tuning » Transformer
