There are about 9200-10000 known species of birds in the world. Many of the species have gone extinct and many are in grave danger. In this porject I am classifying 275 species of birds
I have built a simple classifier using a pre-trained Mobile network V2 to classify 275 different species of birds. I have used Keras (an open source neural network library) to build this classifier. The model (bird_model.ipynb) contains a pre-trained Xception network.
I've got around 96% accuracy in the Test set.
Data set of 275 bird species. Training Images : 39364 , Validation Images : 1375 , Testing Images :1375
All images are 224 X 224 X 3 color images in jpg format.
Datatset Source : https://www.kaggle.com/gpiosenka/100-bird-species