This repository contains model and training notebook for traffic sign recognition
The dataset used for training is German Traffic Sign Recognition Benchmark (GTSRB) containing 43 classes of traffic signs. The training set contains 39209 labeled images and the test set contains 12630 unlabelled images.
Using and YOLOv5 for training and achieved >93% mAP on the dataset.
Used WandB to log hyperparameters and output metrics from runs.
Download the model