License-plate-Detection

Transfer Learning with imagenet models for Vehicle and License plate detection task.

Run the generate_dataset.py to generate cropped images with respective transformations

The code crops the original 1980 x 1080 into 512x512 images and transforms the annotations for the same.

Dataset used is the UFPR - ALPR dataset. (Dataset is not present in the repo)

Implementing a YOLO like architecture. Model with best metrics in model directory.

Documentation in the code and the readme will be updated as the project is in progress.