All projects

Trajectory Planning for Autonomous Vehicles

Reinforcement-learning-based driving in the CARLA simulator.

Dec 2019 Machine Learning

An end-to-end autonomous-driving agent trained and deployed in the CARLA simulator.

  • Implemented and deployed an autonomous-driving model in the CARLA simulator.
  • Used FC-DenseNet56 and PSPNet for semantic segmentation of live road video.
  • Employed Deep Q-Networks for reinforcement-learning-based path planning.
Reinforcement LearningDeep Q-NetworksCARLASemantic Segmentation