End-to-End Autonomous Driving
End-to-end driving models that learn directly from sensor inputs to control commands, including transformer-based, imitation learning and reinforcement learning approaches.
Autonomous driving research topics
BEV Perception
Bird's Eye View (BEV) perception for autonomous driving — camera-to-BEV transformations, LiDAR BEV, multi-sensor BEV fusion and temporal BEV modeling.
3D Object Detection
3D object detection for autonomous driving using LiDAR, cameras, or sensor fusion — 3D bounding boxes, object classification, velocity estimation and tracking.
Motion Prediction
Motion prediction and trajectory forecasting for autonomous driving — predicting future positions of vehicles, pedestrians and cyclists in complex traffic scenarios.
Path Planning
Path planning and motion planning for autonomous vehicles — route planning, trajectory optimization, decision-making and planning in complex urban traffic.
LiDAR Perception
LiDAR-based 3D perception for autonomous driving — point cloud processing, 3D detection, segmentation, compression and sensor fusion with cameras and radar.
Occupancy Prediction
Occupancy grid and occupancy network methods for autonomous driving — volumetric 3D environment representation for obstacle avoidance and free-space estimation.
Sensor Fusion
Multi-sensor fusion for autonomous driving — combining LiDAR, camera, radar and ultrasonic sensors for robust all-weather environment perception.
Autonomous Driving Simulation
Simulation environments and synthetic data generation for autonomous driving — CARLA, nuPlan, Waymax, IDSim and closed-loop evaluation methodologies.
Autonomous Driving Research
General autonomous driving research covering the full AV stack — perception, prediction, planning and control for self-driving vehicles and ADAS systems.