TrajNet
收藏资源简介:
TrajNet是一种基于轨迹的大规模活动基准,为测试方法提供统一的评估系统。挑战包括预测3,161个人轨迹,根据世界平面坐标 (3.2秒) 观察每个轨迹的8个连续地面真值,即t-7,t-6,…,t,并预测以下12个轨迹 (4.8秒),即t 1,…,t 12。 TrajNet是多个数据集的超集,包括各种基于轨迹的活动预测数据集,例如BIWI酒店,crowdusucy,MOT PETS和斯坦福无人机数据集 (SDD)。
TrajNet is a large-scale trajectory-based activity benchmark that provides a unified evaluation system for testing methods. The challenge entails predicting 3,161 individual trajectories: given 8 consecutive ground truth observations of each trajectory in world plane coordinates over 3.2 seconds (i.e., t-7, t-6, …, t), the task is to predict the subsequent 12 trajectories over 4.8 seconds (i.e., t₁, t₂, …, t₁₂). TrajNet is a superset of multiple datasets, encompassing various trajectory-based activity prediction datasets such as BIWI Hotel, crowdusucy, MOT PETS, and the Stanford Drone Dataset (SDD).




