austin_sirius_dataset_converted_externally_to_rlds

  • Description:

Franka tabletop manipulation tasks

Split Examples
'train' 559
  • Feature structure:
FeaturesDict({
    'episode_metadata': FeaturesDict({
        'file_path': Text(shape=(), dtype=string),
    }),
    'steps': Dataset({
        'action': Tensor(shape=(7,), dtype=float32),
        'action_mode': Tensor(shape=(1,), dtype=float32),
        'discount': Scalar(shape=(), dtype=float32),
        'intv_label': Tensor(shape=(1,), dtype=float32),
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'language_embedding': Tensor(shape=(512,), dtype=float32),
        'language_instruction': Text(shape=(), dtype=string),
        'observation': FeaturesDict({
            'image': Image(shape=(84, 84, 3), dtype=uint8),
            'state': Tensor(shape=(8,), dtype=float32),
            'state_ee': Tensor(shape=(16,), dtype=float32),
            'state_gripper': Tensor(shape=(1,), dtype=float32),
            'state_joint': Tensor(shape=(7,), dtype=float32),
            'wrist_image': Image(shape=(84, 84, 3), dtype=uint8),
        }),
        'reward': Scalar(shape=(), dtype=float32),
    }),
})
  • Feature documentation:
Feature Class Shape Dtype Description
FeaturesDict
episode_metadata FeaturesDict
episode_metadata/file_path Text string Path to the original data file.
steps Dataset
steps/action Tensor (7,) float32 Robot action, consists of [3x ee relative pos, 3x ee relative rotation, 1x gripper action].
steps/action_mode Tensor (1,) float32 Type of interaction. -1: initial human demonstration. 1: intervention. 0: autonomuos robot execution (includes pre-intervention class)
steps/discount Scalar float32 Discount if provided, default to 1.
steps/intv_label Tensor (1,) float32 Same as action_modes, except 15 timesteps preceding intervention are labeled as -10.
steps/is_first Tensor bool
steps/is_last Tensor bool
steps/is_terminal Tensor bool
steps/language_embedding Tensor (512,) float32 Kona language embedding. See https://tfhub.dev/google/universal-sentence-encoder-large/5
steps/language_instruction Text string Language Instruction.
steps/observation FeaturesDict
steps/observation/image Image (84, 84, 3) uint8 Main camera RGB observation.
steps/observation/state Tensor (8,) float32 Default robot state, consists of [7x robot joint state, 1x gripper state].
steps/observation/state_ee Tensor (16,) float32 End-effector state, represented as 4x4 homogeneous transformation matrix of ee pose.
steps/observation/state_gripper Tensor (1,) float32 Robot gripper opening width. Ranges between ~0 (closed) to ~0.077 (open)
steps/observation/state_joint Tensor (7,) float32 Robot 7-dof joint information.
steps/observation/wrist_image Image (84, 84, 3) uint8 Wrist camera RGB observation.
steps/reward Scalar float32 Reward if provided, 1 on final step for demos.
  • Citation:
@inproceedings{liu2022robot,
    title = {Robot Learning on the Job: Human-in-the-Loop Autonomy and Learning During Deployment},
    author = {Huihan Liu and Soroush Nasiriany and Lance Zhang and Zhiyao Bao and Yuke Zhu},
    booktitle = {Robotics: Science and Systems (RSS)},
    year = {2023}
}