Using pre-trained ResNet50v2 model for multiclass image classification of Oxford-IIIT Pets dataset
The Oxford-IIIT Pet Dataset is a 37 category pet dataset with roughly 200 images for each class created by the Visual Geometry Group at Oxford. The images have a large variations in scale, pose and lighting. All images have an associated ground truth annotation of breed, head ROI, and pixel level trimap segmentation. [kaggle.com]
Original dataset was downloaded from https://www.robots.ox.ac.uk/~vgg/data/pets/.
1: Import Libraries and Initial Setup
2: Import Data
3: Understanding the dataset
4: Data Preparation: Image Resizing, Label One-hot Encoding, Image (Data) Augmentation, Batching and Prefetching
5: Model Setup
6: Model Training
7: Model Evaluation
8: Finetune Model
9: Run Model on new data
- [ResNet50V2]
- [InceptionV3]
- [MobileNetV2]
The Application of these models can be found in this repo: [Pet Image Classification Web App]