Convolutional neural network-based image recognition for animals

Abstract

In this paper, the basic of convolutional neural network (CNN) and machine learning (ML) was introduced firstly. With the help of TensorFlow, a model building tool, including how they work together to process an image and extract the basic information of it, 25000 images were used including both cats and dogs to train the CNN, which makes the differences between cats and dogs accessible, therefore identifying the cats and dogs. After training the program with such data sets, this work calculated the probability of correctly identifying the animals in the image and finally analyzed it by using the existing data sets.