Traffic Light Detection and Classification with TensorFlow 🚦

Traffic Light Detection and Classification with TensorFlow, Transfer Learning and COCO Model.

Using both the COCO Model to detect the traffic light and then the Custom Traffic Light Classification Model to classify if the traffic light is either Red, Gren, Yellow or None.

The Traffic Light Classification Model was trained using Transfer Learning from the InceptionV3 Model, and with over 1000 Traffic Light Images, 250 Images for each Traffic Light Class.

Check-it out

The whole Code of the model is in the traffic_lights_detection.ipynb file.

Resources:

Traffic Light Datasets:

• https://www.kaggle.com/datasets/wjybuqi/traffic-light-detection-dataset

• https://www.kaggle.com/datasets/sachsene/carla-traffic-lights-images

Traffic Light Classification Model:

• https://drive.google.com/file/d/17cbVavxS0xmDFvZm7bn7mVqS3BAUuSN8/view?usp=sharing

Code:

• https://github.com/mendez-luisjose/Traffic-Light-Detection-and-Clasification-with-TensorFlow