VTP-TL

Vtp-tl中的数据来自带有交通信号灯的城市十字路口，用于预测一天中不同时间的车辆轨迹，并提供广泛的实际驾驶场景。我们使用无人机在交通路口上方70至120米处悬停，尽可能静态地记录在非高峰时段、高峰时段和晚上白天鸟瞰通过该区域的车辆轨迹。

The data in Vtp-tl is collected from urban intersections equipped with traffic signals, serving the task of predicting vehicle trajectories at different times of a day and covering a wide range of real-world driving scenarios. We deployed drones hovering 70 to 120 meters above these intersections to stably record bird's-eye view vehicle trajectories passing through the area during off-peak hours, peak hours, daytime and nighttime as much as possible.

搜集汇总

背景与挑战

背景概述

VTP-TL是一个用于车辆轨迹预测的数据集，包含城市十字路口在不同时间段的交通数据，由无人机拍摄并覆盖多种驾驶场景，由郑州大学于2022年发布。