Techniques for determining output from multiple sensor modalities are discussed herein. Features from a radar sensor, a lidar sensor, and an image sensor may be input into respective models to determine respective intermediate outputs and associated confidence levels associated with tracking associated with the object. Intermediate outputs from the radar model, the lidar model, and the vision model may be input into a fusion model to determine a fusion confidence level and a fusion output associated with tracking. The fused confidence level and each confidence level are compared to a threshold to generate a track for transmission to a planning system or a predictive system of the autonomous vehicle. Additionally, the vehicle controller may control the autonomous vehicle based on the tracking and/or confidence level.
本文讨论了用于确定来自多个传感器模态的输出的技术。可以将来自雷达传感器、激光雷达传感器和图像传感器的特征输入到相应的模型中,以确定与对象相关联的跟踪相关联的相应中间输出和关联的置信度水平。可以将来自雷达模型、激光雷达模型和视觉模型的中间输出输入到融合模型中以确定与跟踪相关联的融合置信度水平和融合输出。将融合置信度水平和各个置信度水平与阈值进行比较以生成跟踪用以传输到自主车辆的规划系统或者预测系统。另外地,车辆控制器可以基于跟踪和/或置信度水平来控制自主车辆。
Tracking confidence model
跟踪置信度模型
26.07.2024
Patent
Elektronische Ressource
Chinesisch
IPC: | B60W CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION , Gemeinsame Steuerung oder Regelung von Fahrzeug-Unteraggregaten verschiedenen Typs oder verschiedener Funktion |
Combined tracking confidence and classification model
Europäisches Patentamt | 2022
|Multi-target tracking on confidence maps: An application to people tracking
British Library Online Contents | 2013
|