Research: Autonomous Driving

Traffic sign identification using camera images from vehicles is key towards autonomous path planning and driving. However, weather, lighting variations and vandalism can lead to misclassifications. In this work, we present an end-to-end framework to augment traffic sign training data using optimal reinforcement learning policies and a variety of Generative Adversarial Network models that can then be used to train traffic sign detector modules.

Annotated Data for Reinforcement Learning policies for auto-augment

https://drive.google.com/drive/folders/1KEWpPejxgT9Ovno72iFdsmJ-pMbnvIPj?usp=sharing

Annotated Data for GAN implementation (Day to night conversion)

https://drive.google.com/drive/u/0/folders/1TjhUHtJbYjYExDS0FOEo59lmoBiirkth