The goal of this project is to distinguish images that contain an image of car plate.
- Trained multiple deep CNN models to classify images that contain car plates.
- Coded relatively complex deep models using Keras with caching support.
- Code for implementing models (Xception, ResNet, MobileNet) is places in
notebooksdirectory. Scriptsdirectory contains codes for a custom data augmentation method and running and evaluating the saved models.- This project is done in collaboration with Mahsa Anvarian as part of our Computer Vision course.
- The best project of the class, both in terms of execution time and F1 score.