{"slug": "two2four-generative-quadruped-puppeteering-from-human-motion", "title": "Two2Four: Generative Quadruped Puppeteering from Human Motion", "summary": "Researchers from DisneyResearch|Studios and ETH Zurich have developed Two2Four, a generative framework that automatically converts ordinary human motion into plausible and controllable quadruped animations, using a two-stage diffusion model trained on quadruped motion data. The system supports actions such as walking, running, jumping, sitting, and lying, and offers fine-grained control like head movement and individual limb puppeteering, outperforming existing retargeting methods in realism and controllability for animation and virtual production.", "body_md": "# Two2Four: Generative Quadruped Puppeteering from Human Motion\n\n### In this work, we present an automatic human-to-quadruped puppeteering framework that produces plausible and controllable quadruped motions from ordinary human motion data.\n\n**July 30, 2026**\n\n**arXiv (2026)**\n\n#### Authors\n\nFatemeh Zargarbashi (DisneyResearch|Studios/ETH Zurich)\n\nZehong Qiu (DisneyResearch|Studios)\n\nDhruv Agrawal (DisneyResearch|Studios/ETH Zurich)\n\nStelian Coros (ETH Zurich)\n\nRobert W. Sumner (DisneyResearch|Studios/ETH Zurich)\n\nMartin Guay (DisneyResearch|Studios)\n\nJakob Buhmann (DisneyResearch|Studios)\n\n#### Two2Four: Generative Quadruped Puppeteering from Human Motion\n\nRealistic animal motion for virtual production is typically obtained either through motion capture of highly trained performers who accurately mimic animal behavior, or by retargeting ordinary human motion using complex control setups. Both approaches are challenging and often fail to fully reproduce the nuances of natural animal motion, motivating data-driven alternatives. We present an automatic human-to-quadruped puppeteering framework that produces plausible and controllable quadruped motions from ordinary human motion data. Our approach employs a two-stage generative diffusion model trained purely on quadruped motion data. By introducing a structured conditioning and inpainting strategy, our method supports a wide range of actions, including walking, running, jumping, sitting, and lying. Furthermore, we enable fine-grained intuitive control of the quadruped motion such as head movement control and individual limb puppeteering. Experimental results demonstrate improved motion realism and controllability compared to existing retargeting approaches, highlighting the effectiveness of our framework as a tool for animation and virtual production applications.", "url": "https://wpnews.pro/news/two2four-generative-quadruped-puppeteering-from-human-motion", "canonical_source": "https://studios.disneyresearch.com/2026/07/30/two2four-generative-quadruped-puppeteering-from-human-motion/", "published_at": "2026-07-30 17:32:18+00:00", "updated_at": "2026-08-19 04:15:01.671444+00:00", "lang": "en", "topics": ["generative-ai", "artificial-intelligence", "machine-learning"], "entities": ["DisneyResearch|Studios", "ETH Zurich", "Fatemeh Zargarbashi", "Zehong Qiu", "Dhruv Agrawal", "Stelian Coros", "Robert W. Sumner", "Martin Guay"], "alternates": {"html": "https://wpnews.pro/news/two2four-generative-quadruped-puppeteering-from-human-motion", "markdown": "https://wpnews.pro/news/two2four-generative-quadruped-puppeteering-from-human-motion.md", "text": "https://wpnews.pro/news/two2four-generative-quadruped-puppeteering-from-human-motion.txt", "jsonld": "https://wpnews.pro/news/two2four-generative-quadruped-puppeteering-from-human-motion.jsonld"}}