cmu_franka_exploration_dataset_converted_externally_to_rlds

  • 説明

おもちゃのキッチンを探検するフランカ

スプリット
'train' 199
  • 機能の構造:
FeaturesDict({
    'episode_metadata': FeaturesDict({
        'file_path': Text(shape=(), dtype=string),
    }),
    'steps': Dataset({
        'action': Tensor(shape=(8,), dtype=float32),
        'discount': Scalar(shape=(), 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({
            'highres_image': Image(shape=(480, 640, 3), dtype=uint8),
            'image': Image(shape=(64, 64, 3), dtype=uint8),
        }),
        'reward': Scalar(shape=(), dtype=float32),
        'structured_action': Tensor(shape=(8,), dtype=float32),
    }),
})
  • 機能ドキュメント:
特徴クラスDタイプ説明
特徴辞書
エピソード_メタデータ特徴辞書
エピソードメタデータ/ファイルパス文章元のデータ ファイルへのパス。
ステップデータセット
ステップ/アクションテンソル(8,) float32ロボット アクションは、[エンド エフェクターの位置 3x、エンド エフェクターの方向 3x、グリッパー アクション 1x、エピソードの終了 1x] で構成されます。
歩数/割引スカラーfloat32割引が指定されている場合、デフォルトは 1 です。
ステップ/is_firstテンソルブール
ステップ/is_lastテンソルブール
ステップ/is_terminalテンソルブール
ステップ/言語_埋め込みテンソル(512,) float32コナ言語の埋め込み。 https://tfhub.dev/google/universal-sentence-encoder-large/5を参照してください。
ステップ/言語説明文章言語指導。
ステップ/観察特徴辞書
ステップ/観察/highres_image画像(480、640、3) uint8高解像度メインカメラ観察
手順・観察・イメージ画像(64、64、3) uint8メインカメラRGB観察。
歩数/報酬スカラーfloat32提供されている場合は報酬、デモの最終ステップで 1。
ステップ/構造化アクションテンソル(8,) float32ハイブリッド アフォーダンスとエンドエフェクター制御で構成される構造化アクション。「ヒューマン ビデオからの構造化世界モデル」で説明されています。
  • 引用
@inproceedings{mendonca2023structured,
              title={Structured World Models from Human Videos},
              author={Mendonca, Russell  and Bahl, Shikhar and Pathak, Deepak},
              journal={RSS},
              year={2023}
            }