Genie Envisioner: A Unified World Foundation Platform for Robotic Manipulation
National University of Singapore · Fudan University · Shanghai Jiaotong University · beihang university · Agibot · agibot · Harbin Institute of Technology · Beihang University · University of California, Berkeley · AgiBot
PDF 由论文原始站点提供，PaperCompass 不保存论文文件。
摘要
We introduce Genie Envisioner (GE), a unified world foundation platform for robotic manipulation that jointly learns visual representations and action policies within a single video-generative framework. At its core, GE-Base is a large-scale instruction-conditioned video diffusion model that captures the spatial, temporal, and semantic dynamics of real-world robotic interactions in a structured latent space. Building on this foundation, GE-Act employs a lightweight flow-matching decoder to map latent representations into executable action trajectories, enabling precise and generalizable policy inference across diverse embodiments with minimal supervision. Trained on over 1 million manipulation episodes, GE supports both short- and long-horizon tasks, and generalizes across embodiments. All code, models, and benchmarks will be released publicly.