시파르100

이 데이터 세트는 각각 600개의 이미지를 포함하는 100개의 클래스가 있다는 점을 제외하면 CIFAR-10과 같습니다. 클래스당 500개의 학습 이미지와 100개의 테스트 이미지가 있습니다. CIFAR-100의 100개 클래스는 20개의 수퍼클래스로 그룹화됩니다. 각 이미지에는 "fine" 레이블(이미지가 속한 클래스)과 "coarse" 레이블(이미지가 속한 수퍼 클래스)이 있습니다.

나뉘다
'test' 10,000
'train' 50,000
  • 기능 구조 :
FeaturesDict({
    'coarse_label': ClassLabel(shape=(), dtype=int64, num_classes=20),
    'id': Text(shape=(), dtype=string),
    'image': Image(shape=(32, 32, 3), dtype=uint8),
    'label': ClassLabel(shape=(), dtype=int64, num_classes=100),
})
  • 기능 문서 :
특징 수업 모양 D타입 설명
풍모Dict
거친_라벨 클래스 레이블 int64
ID 텍스트
영상 영상 (32, 32, 3) uint8
상표 클래스 레이블 int64

심상

  • 인용 :
@TECHREPORT{Krizhevsky09learningmultiple,
    author = {Alex Krizhevsky},
    title = {Learning multiple layers of features from tiny images},
    institution = {},
    year = {2009}
}
,

이 데이터 세트는 각각 600개의 이미지를 포함하는 100개의 클래스가 있다는 점을 제외하면 CIFAR-10과 같습니다. 클래스당 500개의 학습 이미지와 100개의 테스트 이미지가 있습니다. CIFAR-100의 100개 클래스는 20개의 수퍼클래스로 그룹화됩니다. 각 이미지에는 "fine" 레이블(이미지가 속한 클래스)과 "coarse" 레이블(이미지가 속한 수퍼 클래스)이 있습니다.

나뉘다
'test' 10,000
'train' 50,000
  • 기능 구조 :
FeaturesDict({
    'coarse_label': ClassLabel(shape=(), dtype=int64, num_classes=20),
    'id': Text(shape=(), dtype=string),
    'image': Image(shape=(32, 32, 3), dtype=uint8),
    'label': ClassLabel(shape=(), dtype=int64, num_classes=100),
})
  • 기능 문서 :
특징 수업 모양 D타입 설명
풍모Dict
거친_라벨 클래스 레이블 int64
ID 텍스트
영상 영상 (32, 32, 3) uint8
상표 클래스 레이블 int64

심상

  • 인용 :
@TECHREPORT{Krizhevsky09learningmultiple,
    author = {Alex Krizhevsky},
    title = {Learning multiple layers of features from tiny images},
    institution = {},
    year = {2009}
}