Traffic Sign Recognition Dataset and Data Augmentation

Although there are many datasets for traffic sign classification, there are few datasets collected for traffic sign recognition and few of them obtain enough instances especially for training a model with the deep learning method. The deep learning method is almost the only way to train a model for real-world usage that covers various highly similar classes compared with the traditional way such as through color, shape, etc. Plus, due to the appearance frequency of different classes of traffic signs in the real world, the imbalance between different classes' instances in the datasets makes the training results even worse. Also, for some certain sign classes, their sign meanings were destined to can't get enough instances in the dataset. To solve this problem, we purpose a unique data augmentation method for the traffic sign recognition dataset that takes advantage of the standard of the traffic sign. We called it TSR dataset augmentation. We based on the benchmark Tsinghua-Tencent 100K (TT100K) dataset to verify the unique data augmentation method. The iteration version datasets based on TT100K, data augmentation method source code and the training results introduced in this paper are publicly available. Deep learning is a machine learning technique that teaches computers to do what humans are born with: learn by example. In deep learning, computer models learn to perform tasks directly from images, text, or sound. Deep learning models can achieve state-of-the-art precision, sometimes exceeding human levels. Models are trained by using a large set of labeled data[1].