Dog-Breed-Identification-Python

This repo contains all the python code for the dog breed classification using CNNs.

How to run these files:

1. Resnet50

1. Upload the notebook to google colab
2. Mount your drive. Upload the data that is present in kaggle to 'dog_breed_identification_files' folder in your drive.
3. Turn on the GPU from the colab settings page.
4. Run the notebook.

2. Inception

1. Upload the notebook to google colab
2. Add the kaggle dog breed into the notebook. The option to add the dataset will be on the panel on the right side of the screen.
3. Mount your drive. Upload the data that is present in kaggle to 'dog_breed_identification_files' folder in your drive.
4. Run the notebook

3. Efficientnet

1. upload the notebook to kaggle
2. Add the kaggle dog breed into the notebook. The option to add the dataset will be on the panel on the right side of the screen.
3. Add the GPU as T4 or p100 on the right side panel.
4. Run the notebook

4. VGG16

1. upload the notebook to kaggle
2. Add the kaggle dog breed into the notebook. The option to add the dataset will be on the panel on the right side of the screen.
3. Add the GPU as T4 or p100 on the right side panel.
4. Run the notebook

5. MyCNN

1. Upload the notebook to google colab
2. Mount your drive. Upload the data that is present in kaggle to 'dog_breed_identification_files' folder in your drive.
3. Turn on the GPU from the colab settings page.
4. Run the notebook.

How to run the Android App

1. Download/Clone the github repo from here
2. Download Android Studio and Install an emulator on Android Studio
3. Open the cloned code/downloaded code using File -> Open
4. Once you open the folder, Android Studio automatically builds the code for you. Wait for it to finish.
5. Once finished, click the play button that is on the top of the screen and wait for the emulator to start.
6. Once the emulator boots up the app will be installed into the emulator.
7. Once the app is opened you can either select/capture the image and then click on predict dog breed button at the bottom.
8. Wait for a few seconds and you will see the results of the model.
9. You will see the top5 predictions in the app.