IDENTIFICATION OF BIRD SPECIES USING CONVOLUTION NUERAL NETWORKS- vGG16

IDENTIFICATION OF BIRD SPECIES USING CONVOLUTION NUERAL NETWORKS- vGG16.

Satvik Srivastav

Department of Computer Science and Engineering

SRM Institute of Science and Technology

Chennai, India ss5465@srmist.edu.in.

Dr. N. Snehalatha

Department of Computer Science and Engineering

SRM Institute of Science and Technology

Chennai, India @srmist.edu.in

Abstract— There are more than 9000 different kinds of birds in our world. Some bird species are hard to find, and if they are found, it's hard to tell what will happen next. To solve this problem, we have an easy and effective way to tell what kind of bird it is based on how it looks. Also, it's easier to understand how people can recognize birds from their pictures than from their sounds. So, we've used something called Convolutional Neural Networks (CNN). CNNs are a strong mix of machine learning techniques that have been shown to work well in image processing. In this paper, a CNN system for classifying bird species is shown. It is trained and tested on the BIRDS 500 SPECIES- IMAGE CLASSIFICATION dataset. By making this dataset and using the algorithm for comparing similarity, this system has been shown to work well in the real world. Using this method, anyone can easily figure out the name of the bird they want to know.
Keywords— Bird species, Machine Learning, Convolutional Neural Networks