Red Round Sign Detection (and Classification)

Setup

1. Setup the environment : Run` pip install -r requirements.txt` to set up env

Download Pre-Process Data

1. Run sh setup.sh to download the data
2. Download the gtsdb tfrecords created from here
3. Run convert_format.py to create the test files for the notebook

Training

The code uses Tensorflow Object Detection API (which is included in the repo), the instructions to run code can be found here

References

1. https://github.com/tensorflow/models/tree/master/research/object_detection
2. https://github.com/aarcosg/traffic-sign-detection