KLASIFIKACIJA SLIKA PRIMENOM FASTAI BIBLIOTEKE
DOI:
https://doi.org/10.24867/08BE13AnticKljučne reči:
fastai, CNN, ResNet, learning rate, confusion matrixApstrakt
U ovom radu je razmatran problem klasifikacije slika primenom konvulicionih mreža realizovanih u sklopu fastai biblioteke. Trebalo je prikupiti podatke, obučiti model i prikazati rezultate.
Reference
[1] Image Classification. Dostupno na:
https://www.sciencedirect.com/topics/computer-science/image-classification
[2] Rethinking the Inception Architecture for Computer Vision. Dostupno na:
https://arxiv.org/abs/1512.00567/
[3] Dictionary of computer vision and image processing.
[4] About Train, Validation and Test Sets in Machine Learning. Dostupno na:
https://towardsdatascience.com/train-validation-and-test-sets-72cb40cba9e7
https://www.sciencedirect.com/topics/computer-science/image-classification
[2] Rethinking the Inception Architecture for Computer Vision. Dostupno na:
https://arxiv.org/abs/1512.00567/
[3] Dictionary of computer vision and image processing.
[4] About Train, Validation and Test Sets in Machine Learning. Dostupno na:
https://towardsdatascience.com/train-validation-and-test-sets-72cb40cba9e7
##submission.downloads##
Objavljeno
2020-05-26
Broj časopisa
Rubrika
Elektrotehničko i računarsko inženjerstvo
