Description
The role is responsible for optimizing perception-model outputs for parking, object detection, and occupancy-grid results; maintaining stable target tracking, handling occlusion and target loss, and constructing drivable areas. It also involves joint debugging with the planning and control module, optimizing post-processing performance and algorithm efficiency for embedded platforms, and defining standards and interfaces. The position requires a bachelor’s degree or higher in a relevant field, C++ programming, and computer vision knowledge, with mass-production parking and multi-sensor fusion experience preferred.
