Görselleştirme : Verilerinizi
Açıklama :
Arabalar veri kümesi, 196 araba sınıfının 16.185 görüntüsünü içerir. Veriler 8.144 eğitim görüntüsüne ve 8.041 test görüntüsüne bölünür, burada her bir sınıf kabaca 50-50'ye bölünür. Sınıflar tipik olarak Marka, Model, Yıl düzeyindedir, örneğin 2012 Tesla Model S veya 2012 BMW M3 coupe.
Ana sayfa : https://ai.stanford.edu/~jkrause/cars/car_dataset.html
Kaynak kodu :
tfds.image_classification.Cars196
Sürümler :
-
2.0.0
: İlk sürüm -
2.0.1
: Web sitesi URL güncellemesi -
2.1.0
(varsayılan): Hata düzeltme https://github.com/tensorflow/datasets/issues/3927
-
İndirme boyutu :
1.82 GiB
Veri kümesi boyutu :
1.85 GiB
Otomatik önbelleğe alındı ( belgeler ): Hayır
Bölmeler :
Bölmek | Örnekler |
---|---|
'test' | 8.041 |
'train' | 8.144 |
- Özellik yapısı :
FeaturesDict({
'bbox': BBoxFeature(shape=(4,), dtype=tf.float32),
'id': Text(shape=(), dtype=tf.string),
'image': Image(shape=(None, None, 3), dtype=tf.uint8),
'label': ClassLabel(shape=(), dtype=tf.int64, num_classes=196),
})
- Özellik belgeleri :
Özellik | Sınıf | Şekil | Dtype | Tanım |
---|---|---|---|---|
ÖzelliklerDict | ||||
bbbox | BBoxÖzelliği | (4,) | tf.float32 | |
İD | Metin | tf.string | ||
resim | resim | (Yok, Yok, 3) | tf.uint8 | |
etiket | sınıf etiketi | tf.int64 |
Denetlenen anahtarlar (bkz
as_supervised
doc ):('image', 'label')
Şekil ( tfds.show_examples ):
- Örnekler ( tfds.as_dataframe ):
- alıntı :
@inproceedings{KrauseStarkDengFei-Fei_3DRR2013,
title = {3D Object Representations for Fine-Grained Categorization},
booktitle = {4th International IEEE Workshop on 3D Representation and Recognition (3dRR-13)},
year = {2013},
address = {Sydney, Australia},
author = {Jonathan Krause and Michael Stark and Jia Deng and Li Fei-Fei}
}