स्मॉलनोर्ब

यह डेटाबेस आकृति से 3डी वस्तु पहचान में प्रयोगों के लिए अभिप्रेत है। इसमें 5 सामान्य श्रेणियों से संबंधित 50 खिलौनों की छवियां हैं: चार पैर वाले जानवर, मानव आकृतियां, हवाई जहाज, ट्रक और कार। वस्तुओं को दो कैमरों द्वारा 6 प्रकाश स्थितियों, 9 उन्नयन (30 से 70 डिग्री प्रत्येक 5 डिग्री), और 18 दिगंश (0 से 340 प्रत्येक 20 डिग्री) के तहत चित्रित किया गया था।

प्रशिक्षण सेट में प्रत्येक श्रेणी के 5 उदाहरण (उदाहरण 4, 6, 7, 8 और 9), और शेष 5 उदाहरणों के परीक्षण सेट (उदाहरण 0, 1, 2, 3 और 5) शामिल हैं।

विभाजित करना उदाहरण
'test' 24,300
'train' 24,300
  • फ़ीचर संरचना :
FeaturesDict({
    'image': Image(shape=(96, 96, 1), dtype=uint8),
    'image2': Image(shape=(96, 96, 1), dtype=uint8),
    'instance': ClassLabel(shape=(), dtype=int64, num_classes=10),
    'label_azimuth': ClassLabel(shape=(), dtype=int64, num_classes=18),
    'label_category': ClassLabel(shape=(), dtype=int64, num_classes=5),
    'label_elevation': ClassLabel(shape=(), dtype=int64, num_classes=9),
    'label_lighting': ClassLabel(shape=(), dtype=int64, num_classes=6),
})
  • फ़ीचर दस्तावेज़ीकरण :
विशेषता कक्षा आकार डीटाइप विवरण
विशेषताएं डिक्ट
छवि छवि (96, 96, 1) uint8
image2 छवि (96, 96, 1) uint8
उदाहरण क्लासलेबल int64
label_azimuth क्लासलेबल int64
लेबल_श्रेणी क्लासलेबल int64
लेबल_ऊंचाई क्लासलेबल int64
label_lighting क्लासलेबल int64
  • उद्धरण :
@article{LeCun2004LearningMF,
  title={Learning methods for generic object recognition with invariance to pose and lighting},
  author={Yann LeCun and Fu Jie Huang and L{\'e}on Bottou},
  journal={Proceedings of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition},
  year={2004},
  volume={2},
  pages={II-104 Vol.2}
}