{"@context":"https://schema.org","@type":"CreativeWork","@id":"https://froggit.ai/public/capsules/4f902b06-789d-405e-bf9c-df10914b3986","identifier":"4f902b06-789d-405e-bf9c-df10914b3986","url":"https://froggit.ai/public/capsules/4f902b06-789d-405e-bf9c-df10914b3986","name":"PhyCo: Learning Controllable Physical Priors for Generative Motion","text":"# PhyCo: Learning Controllable Physical Priors for Generative Motion\n\nSource-backed public reference for arXiv:2604.28169.\n\n**Authors:** Sriram Narayanan, Ziyu Jiang, Srinivasa Narasimhan, Manmohan Chandraker\n**Primary source:** https://arxiv.org/abs/2604.28169\n**Published:** 2026-04-30T17:53:03Z\n**Updated:** 2026-04-30T17:53:03Z\n**Categories:** cs.CV, cs.AI, cs.LG\n\n## Abstract Summary\nModern video diffusion models excel at appearance synthesis but still struggle with physical consistency: objects drift, collisions lack realistic rebound, and material responses seldom match their underlying properties. We present PhyCo, a framework that introduces continuous, interpretable, and physically grounded control into video generation. Our approach integrates three key components: (i) a large-scale dataset of over 100K photorealistic simulation videos where friction, restitution, deformation, and force are systematically varied across diverse scenarios; (ii) physics-supervised fine-tuning of a pretrained diffusion model using a ControlNet conditioned on pixel-aligned physical property maps; and (iii) VLM-guided reward optimization, where a fine-tuned vision-language model evaluates generated videos with targeted physics queries and provides differentiable feedback. This combination enables a generative model to produce physically consistent and controllable outputs through variations in physical attributes-without any simulator or geometry reconstruction at inference. On the Physics-IQ benchmark, PhyCo significantly improves physical realism over strong baselines, and human studies confirm clearer and more faithful control over physical attributes. Our results...\n\n## Public Use Notes\n- This capsule summarizes the paper's arXiv metadata and abstract; it is not an independent replication or endorsement of the paper's claims.\n- Use it as a cited research reference for discovery, retrieval, and agent context.\n- For clinical, security, operational, or deployment-sensitive topics","keywords":["cs.CV","cs.AI","cs.LG"],"about":[],"citation":["https://arxiv.org/abs/2604.28169"],"isPartOf":{"@type":"Dataset","name":"Forge Cascade Knowledge Graph","url":"https://froggit.ai"},"publisher":{"@type":"Organization","name":"Forge Cascade","url":"https://froggit.ai"},"dateCreated":"2026-05-01T06:00:03.025000Z","dateModified":"2026-06-19T03:17:48Z","isBasedOn":"https://arxiv.org/abs/2604.28169","additionalProperty":[{"@type":"PropertyValue","name":"trust_level","value":100},{"@type":"PropertyValue","name":"verification_status","value":"sources_verified"},{"@type":"PropertyValue","name":"provenance_status","value":"valid"},{"@type":"PropertyValue","name":"evidence_level","value":"primary_source"}]}