시스코어 시스코어

SI-Score(Synthetic Interventions on Scenes for Robustness Evaluation)는 객체 크기, 위치 및 회전 각도의 변화에 ​​대한 이미지 분류 모델의 견고성을 평가하기 위한 데이터 세트입니다.

SI-SCORE에서는 물체와 배경을 취하고 물체 크기, 위치 및 회전 각도를 체계적으로 변경하여 이러한 요소가 모델 성능에 미치는 영향을 연구할 수 있습니다. 이미지 레이블 공간은 모델을 쉽게 평가할 수 있는 ImageNet 레이블 공간(1k 클래스)입니다.

데이터 세트에 대한 자세한 내용은 https://github.com/google-research/si-score 에서 확인할 수 있습니다.

FeaturesDict({
    'dataset_label': ClassLabel(shape=(), dtype=int64, num_classes=1000),
    'image': Image(shape=(None, None, 3), dtype=uint8),
    'image_id': int64,
    'label': ClassLabel(shape=(), dtype=int64, num_classes=1000),
})
  • 기능 문서 :
특징 수업 모양 D타입 설명
풍모Dict
dataset_label 클래스 레이블 int64
영상 영상 (없음, 없음, 3) uint8
image_id 텐서 int64
상표 클래스 레이블 int64
@misc{djolonga2020robustness,
      title={On Robustness and Transferability of Convolutional Neural Networks},
      author={Josip Djolonga and Jessica Yung and Michael Tschannen and Rob Romijnders and Lucas Beyer and Alexander Kolesnikov and Joan Puigcerver and Matthias Minderer and Alexander D'Amour and Dan Moldovan and Sylvain Gelly and Neil Houlsby and Xiaohua Zhai and Mario Lucic},
      year={2020},
      eprint={2007.08558},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}

siscore/rotation(기본 구성)

  • 구성 설명 : 변동 요인: 회전

  • 다운로드 크기 : 1.40 GiB

  • 데이터세트 크기 : 1.40 GiB

  • 분할 :

나뉘다
'test' 39,540

심상

siscore/크기

  • 구성 설명 : 변동 요인: 크기

  • 다운로드 크기 : 3.25 GiB

  • 데이터세트 크기 : 3.27 GiB

  • 분할 :

나뉘다
'test' 92,884

심상

siscore/위치

  • 구성 설명 : 변동 요인: 위치

  • 다운로드 크기 : 18.21 GiB

  • 데이터세트 크기 : 18.31 GiB

  • 분할 :

나뉘다
'test' 541,548

심상