lvis

  • توضیحات :

LVIS: مجموعه داده ای برای تقسیم بندی نمونه واژگان بزرگ.

  • اسناد اضافی : کاوش در کاغذها با کد

  • صفحه اصلی : https://www.lvisdataset.org/

  • کد منبع : tfds.datasets.lvis.Builder

  • نسخه ها :

    • 1.1.0 : فیلدهای neg_category_ids و not_exhaustive_category_ids اضافه شد.
    • 1.2.0 : نام کلاس ها اضافه شده است.
    • 1.3.0 : اسپلیت مینیوال اضافه شد.
    • 1.4.0 (پیش‌فرض) : ماسک‌های تقسیم‌بندی به تقسیم کوچک اضافه شده است.
  • حجم دانلود : 25.38 GiB

  • حجم مجموعه داده : 23.06 GiB

  • ذخیره خودکار ( اسناد ): خیر

  • تقسیم ها :

تقسیم کنید نمونه ها
'minival' 4809
'test' 19822
'train' 100,170
'validation' 19809
  • ساختار ویژگی :
FeaturesDict({
    'image': Image(shape=(None, None, 3), dtype=uint8),
    'image/id': int64,
    'neg_category_ids': Sequence(ClassLabel(shape=(), dtype=int64, num_classes=1203)),
    'not_exhaustive_category_ids': Sequence(ClassLabel(shape=(), dtype=int64, num_classes=1203)),
    'objects': Sequence({
        'area': int64,
        'bbox': BBoxFeature(shape=(4,), dtype=float32),
        'id': int64,
        'label': ClassLabel(shape=(), dtype=int64, num_classes=1203),
        'segmentation': Image(shape=(None, None, 1), dtype=uint8),
    }),
})
  • مستندات ویژگی :
ویژگی کلاس شکل نوع D توضیحات
FeaturesDict
تصویر تصویر (هیچ، هیچ، 3) uint8
تصویر / شناسه تانسور int64
neg_category_ids دنباله (ClassLabel) (هیچ،) int64
not_exhaustive_category_ids دنباله (ClassLabel) (هیچ،) int64
اشیاء دنباله
اشیاء/منطقه تانسور int64
اشیاء/bbox ویژگی BBox (4،) float32
اشیاء / شناسه تانسور int64
اشیاء/برچسب ClassLabel int64
اشیاء/بخش بندی تصویر (هیچ، هیچ، 1) uint8

تجسم

  • نقل قول :
@inproceedings{gupta2019lvis,
  title={ {LVIS}: A Dataset for Large Vocabulary Instance Segmentation},
  author={Gupta, Agrim and Dollar, Piotr and Girshick, Ross},
  booktitle={Proceedings of the {IEEE} Conference on Computer Vision and Pattern Recognition},
  year={2019}
}