This repository contains the Colab notebook, in which an Machine Learning model for dog breed identification from a given image, and is created using Transfer Learning. The data for training and testing is taken from the Kaggle dog breed identification challenge. The base model used for transfer learning is Imagenet-v2 from TensorFlow Hub. Evaluation is done on the basis of Multiclass Log loss as the metric. Link to the dataset: https://www.kaggle.com/c/dog-breed-identification/data.
SriramaBhat/Dog_Breed_Identification
An ML model trained using Transfer Learning from Imagenet-v2 for identifying various dog breeds from the image.
Jupyter Notebook