Yume1.5：文本控制的交互式世界生成模型
Yume1.5: A Text-Controlled Interactive World Generation Model
统一上下文压缩加线性注意力、以及双向注意力蒸馏，是把扩散视频生成压到实时流式（单卡 A100 上 540p 12fps）的可复用手段，做交互式世界模型或长视频生成的人可以直接借鉴其历史上下文压缩与蒸馏流程。
原文依据 a long-video generation method combining unified context compression and linear attention
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论近期方法已展现出利用扩散模型生成可交互、可探索世界的潜力。然而，这些方法大多面临参数规模过大、依赖冗长的推理步数以及历史上下文快速增长等关键挑战，严重限制了实时性能，并且缺乏文本控制的生成能力。为应对这些挑战，我们提出 Yume1.5，一个旨在从单张图像或文本提示生成逼真、可交互且连续世界的新框架。Yume1.5 通过精心设计的框架实现这一目标，该框架支持基于键盘对生成世界进行探索。该框架包含三个核心组件：（1）一种结合统一上下文压缩与线性注意力的长视频生成方法；（2）一种基于上下文压缩的双向注意力蒸馏方法，配合增强的文本嵌入方案，用于实时流式视频生成。Yume1.5 仅使用单张 A100 GPU，在 540p 分辨率下即可达到平均 12 fps 的生成速度；（3）一种用于生成世界事件的文本控制方法。Github：https://github.com/stdstu12/YUME。
英文原文摘要 1190 字符
Recent approaches have demonstrated the promise of using diffusion models to generate interactive and explorable worlds. However, most of these methods face critical challenges such as excessively large parameter sizes, reliance on lengthy inference steps, and rapidly growing historical context, which severely limit real-time performance and lack text-controlled generation capabilities.To address these challenges, we propose Yume1.5, a novel framework designed to generate realistic, interactive, and continuous worlds from a single image or text prompt. Yume1.5 achieves this through a carefully designed framework that supports keyboard-based exploration of the generated worlds. The framework comprises three core components: (1) a long-video generation method combining unified context compression and linear attention; (2) a context compression-based bidirectional attention distillation approach with an enhanced text embedding scheme for real-time streaming video generation. Yume1.5 achieves an average generation speed of 12 fps at 540p resolution using only a single A100 GPU; (3) a text-controlled method for generating world events. Github: https://github.com/stdstu12/YUME.