austin_sirius_dataset_converted_externally_to_rlds

  • 설명 :

Franka 탁상 조작 작업

나뉘다
'train' 559
  • 기능 구조 :
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),
    }),
})
  • 기능 문서 :
특징 수업 모양 Dtype 설명
특징Dict
에피소드_메타데이터 특징Dict
에피소드_메타데이터/파일_경로 텍스트 원본 데이터 파일의 경로입니다.
단계 데이터세트
단계/작업 텐서 (7,) float32 로봇 동작은 [3x ee 상대 위치, 3x ee 상대 회전, 1x 그리퍼 동작]으로 구성됩니다.
단계/액션_모드 텐서 (1,) float32 상호 작용 유형. -1: 초기 인간 시연. 1: 개입. 0: 자율 로봇 실행(사전 개입 수업 포함)
걸음수/할인 스칼라 float32 할인이 제공되면 기본값은 1입니다.
단계/intv_label 텐서 (1,) float32 개입 이전의 15개 시간 단계가 -10으로 표시된다는 점을 제외하면 action_modes와 동일합니다.
걸음수/is_first 텐서 부울
걸음수/is_last 텐서 부울
단계/is_terminal 텐서 부울
단계/언어_임베딩 텐서 (512,) float32 코나 언어 임베딩. https://tfhub.dev/google/universal-sentence-encoder-large/5를 참조하세요.
단계/언어_지시 텍스트 언어 교육.
단계/관찰 특징Dict
단계/관찰/이미지 영상 (84, 84, 3) uint8 메인 카메라 RGB 관찰.
단계/관찰/상태 텐서 (8,) float32 기본 로봇 상태는 [7x 로봇 관절 상태, 1x 그리퍼 상태]로 구성됩니다.
단계/관찰/state_ee 텐서 (16,) float32 ee 포즈의 4x4 동종 변환 행렬로 표현되는 엔드 이펙터 상태.
단계/관찰/state_gripper 텐서 (1,) float32 로봇 그리퍼 개방 폭. ~0(닫힘) ~ ~0.077(열림) 사이의 범위
단계/관찰/state_joint 텐서 (7,) float32 로봇 7자유도 관절정보.
걸음 수/관찰/wrist_image 영상 (84, 84, 3) uint8 손목 카메라 RGB 관찰.
걸음수/보상 스칼라 float32 제공되는 경우 보상, 데모의 마지막 단계에서 1개.
  • 인용 :
@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}
}