DETERMINING DOG AND CAT BREEDS USING DEEP LEARNING
DOI:
https://doi.org/10.24867/19BE09GrbicKeywords:
image classification, convolutional neural networks, active learning, ResNetAbstract
This paper proposes two approaches to solving the problem of determining dog and cat breeds. The first one is a one-step classification in which the breed is immediately determined. The second one represents the classification in two steps - in the first according to the species, and in the second according to the breed. Both approaches use convolutional neural networks and the second one also uses ResNet50 architecture. In both approaches, the impact of the application of active learning was examined.
References
[1] https://barkhappy.com (приступљено у марту 2022.)
[2] https://www.dokonoko.jp/en (приступљено у марту 2022.)
[3] https://www.catster.com (приступљено у марту 2022.)
[4] Zisserman, Andrea Vedaldi Andrew, and C. V. Jawahar. "Cats and dogs." IEEE conference on computer vision and pattern recognition. 2012.
[5] Cohn, David, Les Atlas, and Richard Ladner. "Improving generalization with active learning." Machine learning 15.2 (1994): 201-221.
[6] Deng, Jia, et al. "Imagenet: A large-scale hierarchical image database." 2009 IEEE conference on computer vision and pattern recognition. Ieee, 2009.
[2] https://www.dokonoko.jp/en (приступљено у марту 2022.)
[3] https://www.catster.com (приступљено у марту 2022.)
[4] Zisserman, Andrea Vedaldi Andrew, and C. V. Jawahar. "Cats and dogs." IEEE conference on computer vision and pattern recognition. 2012.
[5] Cohn, David, Les Atlas, and Richard Ladner. "Improving generalization with active learning." Machine learning 15.2 (1994): 201-221.
[6] Deng, Jia, et al. "Imagenet: A large-scale hierarchical image database." 2009 IEEE conference on computer vision and pattern recognition. Ieee, 2009.
Downloads
Published
2022-09-07
Issue
Section
Electrotechnical and Computer Engineering
