imagenet_sketch

  • Açıklama :

ImageNet-Sketch, 1000 ImageNet sınıfının her biri için 50 adet olmak üzere 50.889 siyah beyaz eskiz görüntüsünden oluşur. Bu resimler orijinal olarak "__ taslağı" için Google Görsel Arama'dan toplanmıştır. 100 görüntü toplandı ve ardından manuel olarak filtrelendi. 50'den az iyi görüntüye sahip sınıflar için, çevirme veya döndürme yoluyla ek görüntüler oluşturuldu.

Bölmek örnekler
'test' 50.889
  • Özellik yapısı :
FeaturesDict({
    'file_name': Text(shape=(), dtype=string),
    'image': Image(shape=(None, None, 3), dtype=uint8),
    'label': ClassLabel(shape=(), dtype=int64, num_classes=1000),
})
  • Özellik belgeleri :
Özellik Sınıf Şekil Dtipi Tanım
ÖzelliklerDict
dosya adı Metin sicim
resim resim (Yok, Yok, 3) uint8
etiket SınıfEtiketi int64

görselleştirme

  • Alıntı :
@inproceedings{wang2019learning,
        title={Learning Robust Global Representations by Penalizing Local Predictive Power},
        author={Wang, Haohan and Ge, Songwei and Lipton, Zachary and Xing, Eric P},
        booktitle={Advances in Neural Information Processing Systems},
        pages={10506--10518},
        year={2019}
}