Deployment of a Dog Breed Classifier Using CNN and ResNet-50 on AWS EC2

This paper presents the development and deployment of a machine learning model for dog breed classification. The model leverages Haar Cascade classifiers for initial detection and Convolutional Neural Networks (CNN) with ResNet-50 for classification. The deployment was carried out on AWS EC2, encountering challenges related to maintaining a live production link with continuous updates. The study discusses model performance, deployment considerations, and solutions to challenges encountered in production.