ARCTIC

该数据集规模较小，包含了精确的3D标签，用于评估在HoloAssist数据集上训练的模型的性能。此外，从HoloAssist进行的零样本迁移学习与在ARCTIC上训练的模型具有竞争力。该数据集在接触序列方面的大小仅为HoloAssist的十分之一。其任务是对现有实验室数据集进行评估，这些数据集具有准确的3D标签。

This is a small-scale dataset containing precise 3D labels, which is designed to evaluate the performance of models trained on the HoloAssist dataset. Moreover, zero-shot transfer learning based on HoloAssist using this dataset delivers competitive performance when compared to models trained on the ARCTIC dataset. In terms of contact sequence scale, the size of this dataset is merely one-tenth that of the HoloAssist dataset. Its core applications also include evaluating models trained on existing laboratory datasets that feature accurate 3D labels.