{
  "name": "RevoTac-OXT",
  "url": [
    "https://huggingface.co/datasets/BrainCoTech/RevoTac-OXT"
  ],
  "version": "0.1.0",
  "description": "Ten scripted simulated tactile manipulation episodes from RevoBench: bimanual Revo3 dexterous hands; 32272 frames at 60 Hz. Rendered Taxim RGB, numeric indentation depth, HydroShear marker displacements, normalized pressure response, and target-specific PhysX pad contact wrench. Simulated data; development trajectories, one episode per task. Stable long holds removed offline without interpolation; original capture frame indices and times retained under edit_provenance. Not physics re-executed. Contributing institutions: HKUST(GZ), BrainCo, Georgia Tech.",
  "type": "dexhand",
  "sensor_image": [
    "RevoBenchSimTaxim"
  ],
  "sensor_array": [
    "RevoBenchSimDepth"
  ],
  "sensor_state": [
    "RevoBenchSimHydroShear",
    "RevoBenchSimPhysXContactWrench",
    "RevoBenchSimPressure"
  ],
  "task": 10,
  "trajectory": 10,
  "frame": 32272,
  "label": [
    "ego"
  ],
  "contact": [
    "davidxulixin@gmail.com"
  ],
  "reference": {
    "paper": "RevoSim: Scalable Multimodal Tactile Simulation for Dexterous Manipulation",
    "website": "https://braincotech.github.io/RevoSim-web/",
    "authors": "Ke Zhang, Lixin Xu, Ziyi Wang, Xiyue Dong, Jie Tan, Lianghu Shao, Xianpeng Wang, Chuanyu Li, Renjing Xu."
  },
  "code_trigger": ""
}