This disclosure discusses techniques for determining a prediction for a top-down representation of an environment based on object motion. A sensor of a first vehicle (such as an autonomous vehicle) may capture sensor data of an environment that may include objects (e.g., a vehicle, pedestrian, bicycle) separate from the first vehicle. A multi-channel image representing a top-down view of an object and environment may be generated based in part on sensor data. Environmental data (object range, speed, lane position, pedestrian crosswalk, etc.) may also be encoded in the image. For example, a plurality of images representing an environment that varies over time may be generated and input into a prediction system configured to output a trajectory template (e.g., a general intent of future motion) and a predicted trajectory (e.g., a more accurate predicted motion) associated with each object. The prediction system may include a machine learning model configured to output a trajectory template and a predicted trajectory.
本公开讨论了用于基于物体运动确定针对环境的自顶向下表示的预测的技术。第一车辆(比如自动驾驶车辆)的传感器可以捕获环境的传感器数据,该环境可以包括与第一车辆分离的物体(例如,车辆、行人、自行车)。用于表示物体和环境的自顶向下视图的多通道图像可以部分地基于传感器数据生成。环境数据(物体范围、速度、车道位置、人行横道等)也可以被编码于图像中。可以生成表示随时间变化的环境的多个图像,并将该多个图像输入预测系统,该预测系统被配置为输出与每个物体相关联的轨迹模板(例如,未来运动的一般意图)和预测轨迹(例如,更为准确的预测运动)。预测系统可以包括被配置为输出轨迹模板和预测轨迹的机器学习模型。
Top-down scene prediction based on object motion
基于物体运动的自顶向下场景预测
2022-08-02
Patent
Electronic Resource
Chinese
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 / G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen |