Autonomous Vehicle Perception: Sensors, Fusion, and Scene Understanding
10-module course on AV perception — camera, LiDAR, radar processing, sensor fusion, and 3D scene reconstruction.
Key Highlights
- ✓ 10 modules with coding labs
- ✓ Camera-based perception (2D detection, segmentation)
- ✓ LiDAR point cloud processing (3D detection)
- ✓ Radar signal processing
- ✓ Multi-sensor fusion (Kalman filter, deep fusion)
- ✓ Real-time inference optimization
Overview
A 10-module course on autonomous vehicle perception. Covers camera, LiDAR, and radar processing, sensor fusion, 3D scene reconstruction, and real-time inference optimization.
What's Inside
Course Modules
(1) AV Perception Overview, (2) Camera Models and Calibration, (3) 2D Detection (YOLO, DETR), (4) Semantic Segmentation, (5) LiDAR Processing, (6) 3D Detection (PointPillars, CenterPoint), (7) Radar Processing, (8) Sensor Fusion, (9) BEV Perception, (10) Deployment Optimization.
Hands-On Labs
Students work with the nuScenes and Waymo Open Dataset. Labs use PyTorch and TensorRT for model training and optimized inference. The capstone builds a multi-sensor fusion pipeline.
Ready to dive in?
Explore this resource and discover more across our 12 technology frontiers.