usc_cloth_sim_converted_externally_to_rlds

  • תיאור :

משימות אינטראקציה מבד פרנקה

לְפַצֵל דוגמאות
'train' 800
'val' 200
  • מבנה תכונה :
FeaturesDict({
    'episode_metadata': FeaturesDict({
        'file_path': Text(shape=(), dtype=string),
    }),
    'steps': Dataset({
        'action': Tensor(shape=(4,), 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({
            'image': Image(shape=(32, 32, 3), dtype=uint8),
        }),
        'reward': Scalar(shape=(), dtype=float32),
    }),
})
  • תיעוד תכונה :
תכונה מעמד צוּרָה Dtype תיאור
FeaturesDict
episode_metadata FeaturesDict
episode_metadata/file_path טֶקסט חוּט נתיב לקובץ הנתונים המקורי.
צעדים מערך נתונים
צעדים/פעולה מוֹתֵחַ (4,) לצוף32 פעולת רובוט, מורכבת ממטרה x,y,z ובורר פקודות<0.5 = פתוח, בוחר>0.5 = סגור.
צעדים/הנחה סקלר לצוף32 הנחה אם ניתנת, ברירת המחדל היא 1.
צעדים/הוא_ראשון מוֹתֵחַ bool
צעדים/הוא_אחרון מוֹתֵחַ bool
steps/is_terminal מוֹתֵחַ bool
שלבים/הטבעת_שפה מוֹתֵחַ (512,) לצוף32 הטבעת שפת Kona. ראה https://tfhub.dev/google/universal-sentence-encoder-large/5
שלבים/הוראת_שפה טֶקסט חוּט הוראת שפה.
צעדים/תצפית FeaturesDict
צעדים/תצפית/תמונה תמונה (32, 32, 3) uint8 תצפית תמונה של בד.
צעדים/פרס סקלר לצוף32 תגמול כמדד ביצועים מנורמל ב-[0, 1].0 = ללא שינוי מהמצב ההתחלתי. 1 = קיפול מושלם. ביצועים פירושם שהבד במצב גרוע יותר מהמצב הראשוני.
@article{salhotra2022dmfd,
    author={Salhotra, Gautam and Liu, I-Chun Arthur and Dominguez-Kuhne, Marcus and Sukhatme, Gaurav S.},
    journal={IEEE Robotics and Automation Letters},
    title={Learning Deformable Object Manipulation From Expert Demonstrations},
    year={2022},
    volume={7},
    number={4},
    pages={8775-8782},
    doi={10.1109/LRA.2022.3187843}
}