robomimic_ph

• 説明：

ロボミミックの熟練した人間のデータセットは、 RoboTurkプラットフォームを使用して 1 人の熟練したオペレーターによって収集されました (トランスポートを除き、2 人の熟練したオペレーターが協力して作業しました)。各データセットは 200 の成功した軌跡で構成されます。

各タスクには 2 つのバージョンがあります。1 つは低次元の観測を含むもの ( low_dim )、もう 1 つは画像を含むもの ( image ) です。

データセットはRLDS 形式に従ってステップとエピソードを表します。

• ホームページ: https://arise-initiative.github.io/robomimic-web/

• ソースコード: tfds.datasets.robomimic_ph.Builder

• バージョン:

  • 1.0.0 : 初期リリース。
  • 1.0.1 (デフォルト): 引用が更新されました。

• 分割:

• 監視キー( as_supervised docを参照): None

• 図( tfds.show_examples ): サポートされていません。

• 例( tfds.as_dataframe ): 欠落しています。

• 引用：

@inproceedings{robomimic2021,
  title={What Matters in Learning from Offline Human Demonstrations for Robot Manipulation},
  author={Ajay Mandlekar and Danfei Xu and Josiah Wong and Soroush Nasiriany
          and Chen Wang and Rohun Kulkarni and Li Fei-Fei and Silvio Savarese
          and Yuke Zhu and Roberto Mart\'{i}n-Mart\'{i}n},
  booktitle={Conference on Robot Learning},
  year={2021}
}

robomimic_ph/lift_ph_image (デフォルト設定)

• ダウンロードサイズ: 798.43 MiB

• データセットのサイズ: 114.47 MiB

• 自動キャッシュ(ドキュメント): はい

• 機能の構造:

FeaturesDict({
    '20_percent': bool,
    '20_percent_train': bool,
    '20_percent_valid': bool,
    '50_percent': bool,
    '50_percent_train': bool,
    '50_percent_valid': bool,
    'episode_id': string,
    'horizon': int32,
    'steps': Dataset({
        'action': Tensor(shape=(7,), dtype=float64),
        'discount': int32,
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'observation': FeaturesDict({
            'agentview_image': Image(shape=(84, 84, 3), dtype=uint8),
            'object': Tensor(shape=(10,), dtype=float64),
            'robot0_eef_pos': Tensor(shape=(3,), dtype=float64, description=End-effector position),
            'robot0_eef_quat': Tensor(shape=(4,), dtype=float64, description=End-effector orientation),
            'robot0_eef_vel_ang': Tensor(shape=(3,), dtype=float64, description=End-effector angular velocity),
            'robot0_eef_vel_lin': Tensor(shape=(3,), dtype=float64, description=End-effector cartesian velocity),
            'robot0_eye_in_hand_image': Image(shape=(84, 84, 3), dtype=uint8),
            'robot0_gripper_qpos': Tensor(shape=(2,), dtype=float64, description=Gripper position),
            'robot0_gripper_qvel': Tensor(shape=(2,), dtype=float64, description=Gripper velocity),
            'robot0_joint_pos': Tensor(shape=(7,), dtype=float64, description=7DOF joint positions),
            'robot0_joint_pos_cos': Tensor(shape=(7,), dtype=float64),
            'robot0_joint_pos_sin': Tensor(shape=(7,), dtype=float64),
            'robot0_joint_vel': Tensor(shape=(7,), dtype=float64, description=7DOF joint velocities),
        }),
        'reward': float64,
        'states': Tensor(shape=(32,), dtype=float64),
    }),
    'train': bool,
    'valid': bool,
})

• 機能ドキュメント:

robomimic_ph/lift_ph_low_dim

• ダウンロードサイズ: 17.69 MiB

• データセットのサイズ: 8.50 MiB

• 自動キャッシュ(ドキュメント): はい

• 機能の構造:

FeaturesDict({
    '20_percent': bool,
    '20_percent_train': bool,
    '20_percent_valid': bool,
    '50_percent': bool,
    '50_percent_train': bool,
    '50_percent_valid': bool,
    'episode_id': string,
    'horizon': int32,
    'steps': Dataset({
        'action': Tensor(shape=(7,), dtype=float64),
        'discount': int32,
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'observation': FeaturesDict({
            'object': Tensor(shape=(10,), dtype=float64),
            'robot0_eef_pos': Tensor(shape=(3,), dtype=float64, description=End-effector position),
            'robot0_eef_quat': Tensor(shape=(4,), dtype=float64, description=End-effector orientation),
            'robot0_eef_vel_ang': Tensor(shape=(3,), dtype=float64, description=End-effector angular velocity),
            'robot0_eef_vel_lin': Tensor(shape=(3,), dtype=float64, description=End-effector cartesian velocity),
            'robot0_gripper_qpos': Tensor(shape=(2,), dtype=float64, description=Gripper position),
            'robot0_gripper_qvel': Tensor(shape=(2,), dtype=float64, description=Gripper velocity),
            'robot0_joint_pos': Tensor(shape=(7,), dtype=float64, description=7DOF joint positions),
            'robot0_joint_pos_cos': Tensor(shape=(7,), dtype=float64),
            'robot0_joint_pos_sin': Tensor(shape=(7,), dtype=float64),
            'robot0_joint_vel': Tensor(shape=(7,), dtype=float64, description=7DOF joint velocities),
        }),
        'reward': float64,
        'states': Tensor(shape=(32,), dtype=float64),
    }),
    'train': bool,
    'valid': bool,
})

• 機能ドキュメント:

robomimic_ph/can_ph_image

• ダウンロードサイズ: 1.87 GiB

• データセットのサイズ: 474.55 MiB

• 自動キャッシュ(ドキュメント): いいえ

• 機能の構造:

FeaturesDict({
    '20_percent': bool,
    '20_percent_train': bool,
    '20_percent_valid': bool,
    '50_percent': bool,
    '50_percent_train': bool,
    '50_percent_valid': bool,
    'episode_id': string,
    'horizon': int32,
    'steps': Dataset({
        'action': Tensor(shape=(7,), dtype=float64),
        'discount': int32,
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'observation': FeaturesDict({
            'agentview_image': Image(shape=(84, 84, 3), dtype=uint8),
            'object': Tensor(shape=(14,), dtype=float64),
            'robot0_eef_pos': Tensor(shape=(3,), dtype=float64, description=End-effector position),
            'robot0_eef_quat': Tensor(shape=(4,), dtype=float64, description=End-effector orientation),
            'robot0_eef_vel_ang': Tensor(shape=(3,), dtype=float64, description=End-effector angular velocity),
            'robot0_eef_vel_lin': Tensor(shape=(3,), dtype=float64, description=End-effector cartesian velocity),
            'robot0_eye_in_hand_image': Image(shape=(84, 84, 3), dtype=uint8),
            'robot0_gripper_qpos': Tensor(shape=(2,), dtype=float64, description=Gripper position),
            'robot0_gripper_qvel': Tensor(shape=(2,), dtype=float64, description=Gripper velocity),
            'robot0_joint_pos': Tensor(shape=(7,), dtype=float64, description=7DOF joint positions),
            'robot0_joint_pos_cos': Tensor(shape=(7,), dtype=float64),
            'robot0_joint_pos_sin': Tensor(shape=(7,), dtype=float64),
            'robot0_joint_vel': Tensor(shape=(7,), dtype=float64, description=7DOF joint velocities),
        }),
        'reward': float64,
        'states': Tensor(shape=(71,), dtype=float64),
    }),
    'train': bool,
    'valid': bool,
})

• 機能ドキュメント:

robomimic_ph/can_ph_low_dim

• ダウンロードサイズ: 43.38 MiB

• データセットのサイズ: 27.73 MiB

• 自動キャッシュ(ドキュメント): はい

• 機能の構造:

FeaturesDict({
    '20_percent': bool,
    '20_percent_train': bool,
    '20_percent_valid': bool,
    '50_percent': bool,
    '50_percent_train': bool,
    '50_percent_valid': bool,
    'episode_id': string,
    'horizon': int32,
    'steps': Dataset({
        'action': Tensor(shape=(7,), dtype=float64),
        'discount': int32,
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'observation': FeaturesDict({
            'object': Tensor(shape=(14,), dtype=float64),
            'robot0_eef_pos': Tensor(shape=(3,), dtype=float64, description=End-effector position),
            'robot0_eef_quat': Tensor(shape=(4,), dtype=float64, description=End-effector orientation),
            'robot0_eef_vel_ang': Tensor(shape=(3,), dtype=float64, description=End-effector angular velocity),
            'robot0_eef_vel_lin': Tensor(shape=(3,), dtype=float64, description=End-effector cartesian velocity),
            'robot0_gripper_qpos': Tensor(shape=(2,), dtype=float64, description=Gripper position),
            'robot0_gripper_qvel': Tensor(shape=(2,), dtype=float64, description=Gripper velocity),
            'robot0_joint_pos': Tensor(shape=(7,), dtype=float64, description=7DOF joint positions),
            'robot0_joint_pos_cos': Tensor(shape=(7,), dtype=float64),
            'robot0_joint_pos_sin': Tensor(shape=(7,), dtype=float64),
            'robot0_joint_vel': Tensor(shape=(7,), dtype=float64, description=7DOF joint velocities),
        }),
        'reward': float64,
        'states': Tensor(shape=(71,), dtype=float64),
    }),
    'train': bool,
    'valid': bool,
})

• 機能ドキュメント:

robomimic_ph/square_ph_image

• ダウンロードサイズ: 2.42 GiB

• データセットのサイズ: 401.28 MiB

• 自動キャッシュ(ドキュメント): いいえ

• 機能の構造:

FeaturesDict({
    '20_percent': bool,
    '20_percent_train': bool,
    '20_percent_valid': bool,
    '50_percent': bool,
    '50_percent_train': bool,
    '50_percent_valid': bool,
    'episode_id': string,
    'horizon': int32,
    'steps': Dataset({
        'action': Tensor(shape=(7,), dtype=float64),
        'discount': int32,
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'observation': FeaturesDict({
            'agentview_image': Image(shape=(84, 84, 3), dtype=uint8),
            'object': Tensor(shape=(14,), dtype=float64),
            'robot0_eef_pos': Tensor(shape=(3,), dtype=float64, description=End-effector position),
            'robot0_eef_quat': Tensor(shape=(4,), dtype=float64, description=End-effector orientation),
            'robot0_eef_vel_ang': Tensor(shape=(3,), dtype=float64, description=End-effector angular velocity),
            'robot0_eef_vel_lin': Tensor(shape=(3,), dtype=float64, description=End-effector cartesian velocity),
            'robot0_eye_in_hand_image': Image(shape=(84, 84, 3), dtype=uint8),
            'robot0_gripper_qpos': Tensor(shape=(2,), dtype=float64, description=Gripper position),
            'robot0_gripper_qvel': Tensor(shape=(2,), dtype=float64, description=Gripper velocity),
            'robot0_joint_pos': Tensor(shape=(7,), dtype=float64, description=7DOF joint positions),
            'robot0_joint_pos_cos': Tensor(shape=(7,), dtype=float64),
            'robot0_joint_pos_sin': Tensor(shape=(7,), dtype=float64),
            'robot0_joint_vel': Tensor(shape=(7,), dtype=float64, description=7DOF joint velocities),
        }),
        'reward': float64,
        'states': Tensor(shape=(45,), dtype=float64),
    }),
    'train': bool,
    'valid': bool,
})

• 機能ドキュメント:

robomimic_ph/square_ph_low_dim

• ダウンロードサイズ: 47.69 MiB

• データセットのサイズ: 29.91 MiB

• 自動キャッシュ(ドキュメント): はい

• 機能の構造:

FeaturesDict({
    '20_percent': bool,
    '20_percent_train': bool,
    '20_percent_valid': bool,
    '50_percent': bool,
    '50_percent_train': bool,
    '50_percent_valid': bool,
    'episode_id': string,
    'horizon': int32,
    'steps': Dataset({
        'action': Tensor(shape=(7,), dtype=float64),
        'discount': int32,
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'observation': FeaturesDict({
            'object': Tensor(shape=(14,), dtype=float64),
            'robot0_eef_pos': Tensor(shape=(3,), dtype=float64, description=End-effector position),
            'robot0_eef_quat': Tensor(shape=(4,), dtype=float64, description=End-effector orientation),
            'robot0_eef_vel_ang': Tensor(shape=(3,), dtype=float64, description=End-effector angular velocity),
            'robot0_eef_vel_lin': Tensor(shape=(3,), dtype=float64, description=End-effector cartesian velocity),
            'robot0_gripper_qpos': Tensor(shape=(2,), dtype=float64, description=Gripper position),
            'robot0_gripper_qvel': Tensor(shape=(2,), dtype=float64, description=Gripper velocity),
            'robot0_joint_pos': Tensor(shape=(7,), dtype=float64, description=7DOF joint positions),
            'robot0_joint_pos_cos': Tensor(shape=(7,), dtype=float64),
            'robot0_joint_pos_sin': Tensor(shape=(7,), dtype=float64),
            'robot0_joint_vel': Tensor(shape=(7,), dtype=float64, description=7DOF joint velocities),
        }),
        'reward': float64,
        'states': Tensor(shape=(45,), dtype=float64),
    }),
    'train': bool,
    'valid': bool,
})

• 機能ドキュメント:

robomimic_ph/transport_ph_image

• ダウンロードサイズ: 15.07 GiB

• データセットのサイズ: 3.64 GiB

• 自動キャッシュ(ドキュメント): いいえ

• 機能の構造:

FeaturesDict({
    '20_percent': bool,
    '20_percent_train': bool,
    '20_percent_valid': bool,
    '50_percent': bool,
    '50_percent_train': bool,
    '50_percent_valid': bool,
    'episode_id': string,
    'horizon': int32,
    'steps': Dataset({
        'action': Tensor(shape=(14,), dtype=float64),
        'discount': int32,
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'observation': FeaturesDict({
            'object': Tensor(shape=(41,), dtype=float64),
            'robot0_eef_pos': Tensor(shape=(3,), dtype=float64, description=End-effector position),
            'robot0_eef_quat': Tensor(shape=(4,), dtype=float64, description=End-effector orientation),
            'robot0_eef_vel_ang': Tensor(shape=(3,), dtype=float64, description=End-effector angular velocity),
            'robot0_eef_vel_lin': Tensor(shape=(3,), dtype=float64, description=End-effector cartesian velocity),
            'robot0_eye_in_hand_image': Image(shape=(84, 84, 3), dtype=uint8),
            'robot0_gripper_qpos': Tensor(shape=(2,), dtype=float64, description=Gripper position),
            'robot0_gripper_qvel': Tensor(shape=(2,), dtype=float64, description=Gripper velocity),
            'robot0_joint_pos': Tensor(shape=(7,), dtype=float64, description=7DOF joint positions),
            'robot0_joint_pos_cos': Tensor(shape=(7,), dtype=float64),
            'robot0_joint_pos_sin': Tensor(shape=(7,), dtype=float64),
            'robot0_joint_vel': Tensor(shape=(7,), dtype=float64, description=7DOF joint velocities),
            'robot1_eef_pos': Tensor(shape=(3,), dtype=float64, description=End-effector position),
            'robot1_eef_quat': Tensor(shape=(4,), dtype=float64, description=End-effector orientation),
            'robot1_eef_vel_ang': Tensor(shape=(3,), dtype=float64, description=End-effector angular velocity),
            'robot1_eef_vel_lin': Tensor(shape=(3,), dtype=float64, description=End-effector cartesian velocity),
            'robot1_eye_in_hand_image': Image(shape=(84, 84, 3), dtype=uint8),
            'robot1_gripper_qpos': Tensor(shape=(2,), dtype=float64, description=Gripper position),
            'robot1_gripper_qvel': Tensor(shape=(2,), dtype=float64, description=Gripper velocity),
            'robot1_joint_pos': Tensor(shape=(7,), dtype=float64, description=7DOF joint positions),
            'robot1_joint_pos_cos': Tensor(shape=(7,), dtype=float64),
            'robot1_joint_pos_sin': Tensor(shape=(7,), dtype=float64),
            'robot1_joint_vel': Tensor(shape=(7,), dtype=float64, description=7DOF joint velocities),
            'shouldercamera0_image': Image(shape=(84, 84, 3), dtype=uint8),
            'shouldercamera1_image': Image(shape=(84, 84, 3), dtype=uint8),
        }),
        'reward': float64,
        'states': Tensor(shape=(115,), dtype=float64),
    }),
    'train': bool,
    'valid': bool,
})

• 機能ドキュメント:

robomimic_ph/transport_ph_low_dim

• ダウンロードサイズ: 294.70 MiB

• データセットのサイズ: 208.05 MiB

• 自動キャッシュ(ドキュメント): shuffle_files=False (トレイン) の場合のみ

• 機能の構造:

FeaturesDict({
    '20_percent': bool,
    '20_percent_train': bool,
    '20_percent_valid': bool,
    '50_percent': bool,
    '50_percent_train': bool,
    '50_percent_valid': bool,
    'episode_id': string,
    'horizon': int32,
    'steps': Dataset({
        'action': Tensor(shape=(14,), dtype=float64),
        'discount': int32,
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'observation': FeaturesDict({
            'object': Tensor(shape=(41,), dtype=float64),
            'robot0_eef_pos': Tensor(shape=(3,), dtype=float64, description=End-effector position),
            'robot0_eef_quat': Tensor(shape=(4,), dtype=float64, description=End-effector orientation),
            'robot0_eef_vel_ang': Tensor(shape=(3,), dtype=float64, description=End-effector angular velocity),
            'robot0_eef_vel_lin': Tensor(shape=(3,), dtype=float64, description=End-effector cartesian velocity),
            'robot0_gripper_qpos': Tensor(shape=(2,), dtype=float64, description=Gripper position),
            'robot0_gripper_qvel': Tensor(shape=(2,), dtype=float64, description=Gripper velocity),
            'robot0_joint_pos': Tensor(shape=(7,), dtype=float64, description=7DOF joint positions),
            'robot0_joint_pos_cos': Tensor(shape=(7,), dtype=float64),
            'robot0_joint_pos_sin': Tensor(shape=(7,), dtype=float64),
            'robot0_joint_vel': Tensor(shape=(7,), dtype=float64, description=7DOF joint velocities),
            'robot1_eef_pos': Tensor(shape=(3,), dtype=float64, description=End-effector position),
            'robot1_eef_quat': Tensor(shape=(4,), dtype=float64, description=End-effector orientation),
            'robot1_eef_vel_ang': Tensor(shape=(3,), dtype=float64, description=End-effector angular velocity),
            'robot1_eef_vel_lin': Tensor(shape=(3,), dtype=float64, description=End-effector cartesian velocity),
            'robot1_gripper_qpos': Tensor(shape=(2,), dtype=float64, description=Gripper position),
            'robot1_gripper_qvel': Tensor(shape=(2,), dtype=float64, description=Gripper velocity),
            'robot1_joint_pos': Tensor(shape=(7,), dtype=float64, description=7DOF joint positions),
            'robot1_joint_pos_cos': Tensor(shape=(7,), dtype=float64),
            'robot1_joint_pos_sin': Tensor(shape=(7,), dtype=float64),
            'robot1_joint_vel': Tensor(shape=(7,), dtype=float64, description=7DOF joint velocities),
        }),
        'reward': float64,
        'states': Tensor(shape=(115,), dtype=float64),
    }),
    'train': bool,
    'valid': bool,
})

• 機能ドキュメント:

robomimic_ph/tool_hang_ph_image

• ダウンロードサイズ: 61.96 GiB

• データセットのサイズ: 9.10 GiB

• 自動キャッシュ(ドキュメント): いいえ

• 機能の構造:

FeaturesDict({
    'episode_id': string,
    'horizon': int32,
    'steps': Dataset({
        'action': Tensor(shape=(7,), dtype=float64),
        'discount': int32,
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'observation': FeaturesDict({
            'object': Tensor(shape=(44,), dtype=float64),
            'robot0_eef_pos': Tensor(shape=(3,), dtype=float64, description=End-effector position),
            'robot0_eef_quat': Tensor(shape=(4,), dtype=float64, description=End-effector orientation),
            'robot0_eef_vel_ang': Tensor(shape=(3,), dtype=float64, description=End-effector angular velocity),
            'robot0_eef_vel_lin': Tensor(shape=(3,), dtype=float64, description=End-effector cartesian velocity),
            'robot0_eye_in_hand_image': Image(shape=(240, 240, 3), dtype=uint8),
            'robot0_gripper_qpos': Tensor(shape=(2,), dtype=float64, description=Gripper position),
            'robot0_gripper_qvel': Tensor(shape=(2,), dtype=float64, description=Gripper velocity),
            'robot0_joint_pos': Tensor(shape=(7,), dtype=float64, description=7DOF joint positions),
            'robot0_joint_pos_cos': Tensor(shape=(7,), dtype=float64),
            'robot0_joint_pos_sin': Tensor(shape=(7,), dtype=float64),
            'robot0_joint_vel': Tensor(shape=(7,), dtype=float64, description=7DOF joint velocities),
            'sideview_image': Image(shape=(240, 240, 3), dtype=uint8),
        }),
        'reward': float64,
        'states': Tensor(shape=(58,), dtype=float64),
    }),
    'train': bool,
    'valid': bool,
})

• 機能ドキュメント:

robomimic_ph/tool_hang_ph_low_dim

• ダウンロードサイズ: 192.29 MiB

• データセットのサイズ: 121.77 MiB

• 自動キャッシュ(ドキュメント): はい

• 機能の構造:

FeaturesDict({
    'episode_id': string,
    'horizon': int32,
    'steps': Dataset({
        'action': Tensor(shape=(7,), dtype=float64),
        'discount': int32,
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'observation': FeaturesDict({
            'object': Tensor(shape=(44,), dtype=float64),
            'robot0_eef_pos': Tensor(shape=(3,), dtype=float64, description=End-effector position),
            'robot0_eef_quat': Tensor(shape=(4,), dtype=float64, description=End-effector orientation),
            'robot0_eef_vel_ang': Tensor(shape=(3,), dtype=float64, description=End-effector angular velocity),
            'robot0_eef_vel_lin': Tensor(shape=(3,), dtype=float64, description=End-effector cartesian velocity),
            'robot0_gripper_qpos': Tensor(shape=(2,), dtype=float64, description=Gripper position),
            'robot0_gripper_qvel': Tensor(shape=(2,), dtype=float64, description=Gripper velocity),
            'robot0_joint_pos': Tensor(shape=(7,), dtype=float64, description=7DOF joint positions),
            'robot0_joint_pos_cos': Tensor(shape=(7,), dtype=float64),
            'robot0_joint_pos_sin': Tensor(shape=(7,), dtype=float64),
            'robot0_joint_vel': Tensor(shape=(7,), dtype=float64, description=7DOF joint velocities),
        }),
        'reward': float64,
        'states': Tensor(shape=(58,), dtype=float64),
    }),
    'train': bool,
    'valid': bool,
})

• 機能ドキュメント: