{
  "title": "HumanEgo: predict actions from hand–object relations",
  "caption": "The author made this diagram from the source figure. Entity tokens contain hand poses in entity frames. The action generator uses RGB and relational context. Dashed auxiliary heads supply training supervision only. The diagram does not show source photographs, prediction plots or task-success measurements.",
  "alt": "The policy flow uses RGB and hand–object tokens to generate bimanual actions, with different prediction heads for training only.",
  "sources": [
    "https://arxiv.org/abs/2605.24934",
    "https://humanego-ai.github.io/"
  ],
  "limitations": [
    "The playback is an example. Playback time is not measured runtime."
  ]
}