A deep feature extractor approach for the recognition of pollen-bearing bees Polen taiyan arilann tanmmasi ifin derin ozellik fikanci bir yaklaim

TÜRKOĞLU M., Uzen H., HANBAY D.

28th Signal Processing and Communications Applications Conference, SIU 2020, Gaziantep, Türkiye, 5 - 07 Ekim 2020, (Tam Metin Bildiri) identifier

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Doi Numarası: 10.1109/siu49456.2020.9302368
  • Basıldığı Şehir: Gaziantep
  • Basıldığı Ülke: Türkiye
  • Anahtar Kelimeler: AlexNet architecture, Convolutional Neural Networks, Object recognition, SVM classifier., VGG16 architecture
  • Samsun Üniversitesi Adresli: Hayır

Özet

In this study, a convolutional neural network (ESA) based feature extracting hybrid model was proposed for the identification of bees carrying pollen or not. The fc6 and fc7 layers of AlexNet and VGG16 which a pre-trained ESA architecture, were used as feature extractors. The performances of the different combinations of the deep properties obtained using the SVM classifier were calculated. The PollenDataset dataset was used to test the proposed model. According to the experimental results, an accuracy score of 97.20% was obtained. As a result, the obtained accuracy score was compared with the state-of-the-art accuracy scores and the proposed model provided better performance than the compared methods.