{
  "id": 5960,
  "url": "https://arxiv.org/abs/2604.11083v2",
  "title": "FlowCoMotion: Text-to-Motion Generation via Token-Latent Flow Modeling",
  "summary": "Text-to-motion generation is driven by learning motion representations for semantic alignment with language. Existing methods rely on either continuous or discrete motion representations. However, continuous representations entangle semantics with dynamics, while discrete representations lose fine-grained motion details. In this context, we propose FlowCoMotion, a novel motion generation framework that unifies both treatments from a modeling perspective. Specifically, FlowCoMotion employs token-",
  "authors": "Dawei Guan, Di Yang, Chengjie Jin, Jiangtao Wang",
  "category": "research",
  "topics": "safety-alignment",
  "orgs": null,
  "regions": null,
  "published_at": "2026-04-13T07:04:47.000Z",
  "fetched_at": "2026-07-14T16:32:06.471Z",
  "source_slug": "arxiv-ethics",
  "source_name": "arXiv",
  "source_homepage": "https://arxiv.org",
  "ethics_ai_record_url": "https://ethics.ai/record/5960",
  "original_url": "https://arxiv.org/abs/2604.11083v2",
  "evidence_status": "source-only",
  "attribution": "via ethics.ai"
}