Mehtap MUTLU, Kevser ÖZDEM, M. Ali AKCAYOL

Derin Öğrenme ile Kuş Türü Sınıflandırma: Karşılaştırmalı Bir Çalışma

Kuş türlerini görüntü üzerinden sınıflandırmaya yönelik çalışmalar hem görüntü içerisindeki renk ve desen çokluğu hem de birbirine çok yakın görsel özelliklere sahip olmalarından dolayı oldukça zordur. Bu çalışmada kuş türlerinin sınıflandırması için altı farklı derin öğrenme modeli uygulanmış ve deneysel sonuçlar kapsamlı bir şekilde karşılaştırılmıştır. Veri kümesi olarak 225 kuş türüne sahip toplam 31316 kuş görüntüsü olan 250 Bird Species isimli veri kümesi kullanılmıştır. Çalışmada 1125 tane görüntü test ve 1125 tane görüntü ise doğrulama için kullanılmı ştır. Veri kümesi üzerinde sırasıyla VGG16, ResNet50, ResNet152V2, InceptionV3, MobileNet ve DenseNet121 derin öğrenme modellerinin doğruluk, kesinlik, hassasiyet ve F1-skoru değerlerine göre karşılaştırması yapılmıştır. Yapılan deneysel çalışmalarda, VGG16 ile %94,6, ResNet50 ile %47,2, ResNet152V2 ile %96,2, InceptionV3 ile %97,5, MobileNet ile %96,9 ve DenseNet121 ile %98,2 doğruluk değerleri elde edilmiştir. En yüksek kesinlik değeri 0,99, hassasiyet değeri 0,99 ve F1-skor değeri 0,99 olarak DenseNet121 ile elde edilmiştir.

Anahtar Kelimeler:

Kuş türü sınıflandırma, VGG16, ResNet50, ResNet152V2, InceptionV3, MobileNet, DenseNet121

Bird Species Classification Using Deep Learning: A Comparative Study

Studies to classify bird species on the basis of images are very difficult due to both the abundance of colors and patterns in the image, and their very close visual characteristics. In this study, six different deep learning models have been applied for the classification of bird species and the experimental results have been compared comprehensively. A dataset named 250 Bird Species, which includes a total of 31316 bird images with 225 bird species, was used as dataset. In the study, 1125 images have been used for the test and 1125 images for the validation. The comparison of VGG16, ResNet50, ResNet152V2, InceptionV3, MobileNet and DenseNet121 deep learning models have been made on the dataset respectively, according to the accuracy, precision, recall and F1-score values. In experimental studies, 94.6% accuracy value has been obtained with VGG16, 47.2% with ResNet50, 96.2% with ResNet152V2, 97.5% with InceptionV3, 96.9% with MobileNet and 98.2% with DenseNet121. DenseNet121 obtained the highest precision value as 0.99, sensitivity value as 0.99 and F1-score value as 0.99.

Keywords:

Bird species classification, VGG16, ResNet50, ResNet152V2, InceptionV3, MobileNet, DenseNet121,

PDF

___

[1] Sangster G., “Integrative taxonomy of birds: the nature and delimitation of species”, Bird Species How They Arise Modify and Vanish, Springer, Cham, (2018).
[2] Gill F. B., “Species taxonomy of birds: which null hypothesis?”, The Auk: Ornithological Advances, 131(2): 150-161, (2014).
[3] Ge Z., McCool C., Sanderson C., Bewley A., Chen Z. and Corke P., “Fine-grained bird species recognition via hierarchical subset learning”, IEEE International Conference on Image Processing (ICIP2015), 561-565, (2015).
[4] Niemi J. and Tanttu J. T., “Deep learning case study for automatic bird identification”, Applied Sciences, 8(11): 2089, (2018).
[5] Liu Y., Sun P., Highsmith M. R., Wergeles N. M., Sartwell J., Raedeke A. and Shang Y., “Performance comparison of deep learning techniques for recognizing birds in aerial images”, IEEE Third International Conference on Data Science in Cyberspace (DSC2018), 317-324, (2018).
[6] Kumar A. and Das S. D., “Bird Species Classification Using Transfer Learning with Multistage Training”, Workshop on Computer Vision Applications, 28-38, (2018).
[7] Huang Y. P. and Basanta H., “Bird image retrieval and recognition using a deep learning platform”, IEEE Access, 7: 66980-66989, (2019).
[8] Hong S. J., Han Y., Kim S. Y., Lee A. Y. and Kim G., “Application of deep-learning methods to bird detection using unmanned aerial vehicle imagery”, Sensors, 19(7): 1651, (2019).
[9] Islam S., Khan S. I. A., Abedin M. M., Habibullah K. M. and Das, A. K., “Bird Species Classification from an Image Using VGG-16 Network”, Proceedings of 7th International Conference on Computer and Communications Management, 38-42, (2019).
[10] Gavali P., Mhetre M. P. A., Patil M. N. C., Bamane M. N. S., and Buva M. H. D., “Bird Species Identification using Deep Learning”, Int. J. Eng. Res. Technol, 8: 68-72, (2019).
[11] Raquel C. R., Alarcon K. M. A. and Figueroa L. L., “Image classification of Philippine bird species using deep learning”, Proceedings of the Workshop on Computation: Theory and Practice (WCTP), 93, (2019).
[12] Ragib K. M., Shithi R. T., Haq S. A., Hasan M., Sakib K. M. and Farah T., “PakhiChini: Automatic Bird Species Identification Using Deep Learning”, Fourth World Conference on Smart Trends in Systems, Security and Sustainability (WorldS4), 1-6, (2020).
[13] Gavali P. and Banu J. S., “Bird Species Identification using Deep Learning on GPU platform”, International Conference on Emerging Trends in Information Technology and Engineering (ic-ETITE, 2020), 1-6, (2020).
[14] Raj S., Garyali S., Kumar S. and Shidnal S., “Image based Bird Species Identification using Convolutional Neural Network”, International Journal of Engineering Research & Technology (IJERT), 9(6): 346, (2020).
[15] Chakraborti T., McCane B., Mills S. and Pal, U., “CoCoNet: A Collaborative Convolutional Network applied to fine-grained bird species classification”, 35th International Conference on Image and Vision Computing New Zealand (IVCNZ), 1-6, (2020).
[16] Trense D. and Tietze D. T., “Studying Speciation: Genomic Essentials and Approaches”, Bird Species How They Arise Modify and Vanish, Springer, Cham, (2018).
[17] Goodfellow I., Bengio Y., Courville A. and Bengio Y., “Deep learning” 1. Cambridge: MIT press, 2, (2016).
[18] Albawi S., Mohammed T. A. and Al-Zawi S., “Understanding of a convolutional neural network”, International Conference on Engineering and Technology (ICET), 1-6, (2017).
[19] Lu J., Behbood V., Hao P., Zuo H., Xue S. and Zhang G., “Transfer learning using computational intelligence: A survey”, Knowledge-Based Systems, 80: 14-23, (2015).
[20] Tan C., Sun F., Kong T., Zhang W., Yang C. and Liu C., “A survey on deep transfer learning”, International conference on artificial neural networks, 270-279, (2018).
[21] https://www.kaggle.com/gpiosenka/100-bird-species/ version/30
[22] Simonyan K. and Zisserman A., “Very deep convolutional networks for large-scale image recognition”, arXiv:1409.1556, (2014).
[23] He K., Zhang X., Ren S. and Sun J., “Deep residual learning for image recognition”, Proceedings of the IEEE conference on computer vision and pattern recognition, 770-778, (2016).
[24] Elshennawy N. M. and Ibrahim D. M., “Deep-Pneumonia Framework Using Deep Learning Models Based on Chest X-Ray Images”, Diagnostics, 10(9): 649, (2020).
[25] Gulli A. and Pal S., “Deep learning with Keras”, Packt Publishing Ltd., 101, (2017).
[26] Howard A. G., Zhu M., Chen B., Kalenichenko D., Wang W., Weyand T. and Adam H., “Mobilenets: Efficient convolutional neural networks for mobile vision applications”, arXiv:1704.04861, (2017).
[27] Huang G., Liu Z., Van Der Maaten L. and Weinberger K. Q., “Densely connected convolutional networks”, Proceedings of the IEEE conference on computer vision and pattern recognition, 4700-4708, (2017).
[28] Coşkun C. and Baykal A., “Veri madenciliğinde sınıflandırma algoritmalarının bir örnek üzerinde karşılaştırılması”, Akademik Bilişim, 1-8, (2011).
[29] Marini A., Facon J. and Koerich A. L., “Bird species classification based on color features”, IEEE International Conference on Systems, Man, and Cybernetics, 4336-4341, (2013).
[30] Lucio D. R., Maldonado Y. and da Costa G., “Bird species classification using spectrograms”, Latin American Computing Conference (CLEI), 1-11, (2015).
[31] Nanni L., Costa Y. M., Lucio D. R., Silla C. N. and Brahnam S., “Combining visual and acoustic features for bird species classification”, IEEE 28th International Conference on Tools with Artificial Intelligence (ICTAI2016), 396-401, (2016).
[32] Pang C., Yao H. and Sun X., “Discriminative features for bird species classification”, Proceedings of International Conference on Internet Multimedia Computing and Service, 256-260, (2014).
[33] Li L., Chen Y., Shen Z., Zhang X., Sang J., Ding Y. and Yu C., “Convolutional neural network for the diagnosis of early gastric cancer based on magnifying narrow band imaging”, Gastric Cancer, 23(1): 126-132, (2020).
[34] Qin Z., Zhang Z., Chen X., Wang C. and Peng, Y., “Fd-mobilenet: Improved mobilenet with a fast downsampling strategy”, 25th IEEE International Conference on Image Processing (ICIP), 1363-1367, (2018).
[35] Rusiecki A., “Trimmed robust loss function for training deep neural networks with label noise”, International Conference on Artificial Intelligence and Soft Computing, 215-222, (2019).