uiuc_d3field

  • বর্ণনা :

অফিস ডেস্ক, বাসনপত্র ইত্যাদি আয়োজন

বিভক্ত উদাহরণ
'train' 192
  • বৈশিষ্ট্য গঠন :
FeaturesDict({
    'episode_metadata': FeaturesDict({
        'file_path': Text(shape=(), dtype=string),
    }),
    'steps': Dataset({
        'action': Tensor(shape=(3,), 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({
            'depth_1': Image(shape=(360, 640, 1), dtype=uint16),
            'depth_2': Image(shape=(360, 640, 1), dtype=uint16),
            'depth_3': Image(shape=(360, 640, 1), dtype=uint16),
            'depth_4': Image(shape=(360, 640, 1), dtype=uint16),
            'image_1': Image(shape=(360, 640, 3), dtype=uint8),
            'image_2': Image(shape=(360, 640, 3), dtype=uint8),
            'image_3': Image(shape=(360, 640, 3), dtype=uint8),
            'image_4': Image(shape=(360, 640, 3), dtype=uint8),
            'state': Tensor(shape=(4, 4), dtype=float32),
        }),
        'reward': Scalar(shape=(), dtype=float32),
    }),
})
  • বৈশিষ্ট্য ডকুমেন্টেশন :
বৈশিষ্ট্য ক্লাস আকৃতি ডিটাইপ বর্ণনা
ফিচারসডিক্ট
episode_metadata ফিচারসডিক্ট
episode_metadata/file_path পাঠ্য স্ট্রিং মূল ডেটা ফাইলের পথ।
পদক্ষেপ ডেটাসেট
পদক্ষেপ/ক্রিয়া টেনসর (৩,) float32 শেষ ফ্রেম থেকে রোবট স্থানচ্যুতি
পদক্ষেপ/ছাড় স্কেলার float32 ডিসকাউন্ট দেওয়া হলে, ডিফল্ট 1.
steps/is_first টেনসর bool
ধাপ/শেষ_শেষ টেনসর bool
steps/is_terminal টেনসর bool
পদক্ষেপ/ভাষা_এম্বেডিং টেনসর (512,) float32 কোন ভাষা এম্বেডিং. https://tfhub.dev/google/universal-sentence-encoder-large/5 দেখুন
পদক্ষেপ/ভাষা_নির্দেশ পাঠ্য স্ট্রিং ভাষার নির্দেশনা।
পদক্ষেপ/পর্যবেক্ষণ ফিচারসডিক্ট
পদক্ষেপ/পর্যবেক্ষণ/গভীরতা_1 ছবি (360, 640, 1) uint16 ক্যামেরা 1 গভীরতা পর্যবেক্ষণ।
পদক্ষেপ/পর্যবেক্ষণ/গভীরতা_2 ছবি (360, 640, 1) uint16 ক্যামেরা 2 গভীরতা পর্যবেক্ষণ।
পদক্ষেপ/পর্যবেক্ষণ/গভীরতা_৩ ছবি (360, 640, 1) uint16 ক্যামেরা 3 গভীরতা পর্যবেক্ষণ।
পদক্ষেপ/পর্যবেক্ষণ/গভীরতা_4 ছবি (360, 640, 1) uint16 ক্যামেরা 4 গভীরতা পর্যবেক্ষণ।
পদক্ষেপ/পর্যবেক্ষণ/চিত্র_1 ছবি (360, 640, 3) uint8 ক্যামেরা 1 আরজিবি পর্যবেক্ষণ।
পদক্ষেপ/পর্যবেক্ষণ/ছবি_২ ছবি (360, 640, 3) uint8 ক্যামেরা 2 আরজিবি পর্যবেক্ষণ।
পদক্ষেপ/পর্যবেক্ষণ/চিত্র_৩ ছবি (360, 640, 3) uint8 ক্যামেরা 3 আরজিবি পর্যবেক্ষণ।
পদক্ষেপ/পর্যবেক্ষণ/ছবি_৪ ছবি (360, 640, 3) uint8 ক্যামেরা 4 আরজিবি পর্যবেক্ষণ।
পদক্ষেপ/পর্যবেক্ষণ/রাষ্ট্র টেনসর (৪, ৪) float32 রোবট শেষ-প্রভাবক অবস্থা
পদক্ষেপ/পুরস্কার স্কেলার float32 প্রদান করা হলে পুরস্কার, ডেমোর জন্য চূড়ান্ত ধাপে 1।
  • তত্ত্বাবধান করা কী (দেখুন as_supervised doc ): None

  • চিত্র ( tfds.show_examples ): সমর্থিত নয়।

  • উদাহরণ ( tfds.as_dataframe ): অনুপস্থিত।

  • উদ্ধৃতি :

@article{wang2023d3field,
  title={D^3Field: Dynamic 3D Descriptor Fields for Generalizable Robotic Manipulation},
  author={Wang, Yixuan and Li, Zhuoran and Zhang, Mingtong and Driggs-Campbell, Katherine and Wu, Jiajun and Fei-Fei, Li and Li, Yunzhu},
  journal={arXiv preprint arXiv:},
  year={2023},
}