Alex Dubranov

Project Details

This project features a machine learning model capable of identifying different dog breeds, created using about 10,000 images from the Kaggle Dog Breed Identification Competition dataset. The model utilizes a convolutional neural network (CNN) with transfer learning, optimized for breed identification, and shows high accuracy in recognizing breeds both from the training set and new images. Along with scripts for training, evaluation, and making predictions, future plans include expanding the dataset and exploring various model architectures to enhance its performance.