FES

该数据集提供了面部检测的事件流和边界框，但遗憾的是，它没有提供眼部注解。此外，FES数据集虽然提供了边界框，但需要手动标注眼部位置，以便进行面部和眼部的追踪任务。

This dataset provides event streams and bounding boxes for face detection, but unfortunately, it does not include eye annotations. Additionally, although the FES Dataset offers bounding boxes, manual annotation of eye positions is required to conduct facial and ocular tracking tasks.