This is a group project created for DATA 2040 Final Project at Brown DSI. Our group name is Code Brew 😎. We will dive deep into the application of YOLOv3 in training our custom detector on traffic signs and apply it in real-time videos.

For more details, you can either read our blog posts (initial ➡️ midway ➡️ final) or watch a 18-min screencast .

If you want to come back to the home page while reading blog posts, just click the Code Brew on the top left corner. If you have any queations, contact us here. Have fun! 🍾

Midway Blog Post

As we have introduced in the initial blog post, the original YOLO v3 model was pre-trained on COCO dataset with 80 dedicated classes. In this project, we will train a custom object detector for traffic signs using Google Colab as our cloud platform. [Read More]