Object Referencing Dataset

该数据集旨在评估IcRegress算法在车外物体参照任务中的性能表现，重点关注了不同驾驶员的特性和各种驾驶条件。数据集包含了多种性能指标，如均方误差（MSE）、平均绝对误差（MAE）、平均相对距离无关误差（MRDE）、到目标的最小距离（MinDT），以及一个名为分段物体（SegObj）的新指标，该指标专门针对物体参照过程中的遮挡问题。此外，该数据集的任务还包括物体参照和手势识别。

This dataset is developed to evaluate the performance of the IcRegress algorithm in out-of-vehicle object reference tasks, with emphasis on the characteristics of diverse drivers and various driving conditions. It incorporates multiple performance metrics, including Mean Squared Error (MSE), Mean Absolute Error (MAE), Mean Relative Distance-independent Error (MRDE), Minimum Distance to Target (MinDT), and a novel metric termed Segmented Object (SegObj), which specifically targets occlusion issues during the object reference process. Furthermore, the tasks covered by this dataset additionally include object reference and gesture recognition.