BioCube

该数据集名为BioCube，是一个多模态、细粒度的全球生态与生物多样性研究数据集，它融合了通过图像、音频记录、环境DNA以及各种环境指标进行的物种观测。数据以Apache Parquet格式存储，这种格式特别优化了输入输出操作。数据集包含了图像和音频的时空匹配，并为每个样本分配了一个独特的标识符以避免重复。该数据集涵盖了超过40,000个物种，旨在支持生物多样性研究和生态系统动态研究任务。

This dataset, named BioCube, is a multimodal, fine-grained global dataset for ecological and biodiversity research. It integrates species observations collected via images, audio recordings, environmental DNA (eDNA) data, and various environmental metrics. Stored in Apache Parquet format, which is specifically optimized for input/output operations, the dataset includes spatiotemporally matched image and audio data, and assigns a unique identifier to each sample to prevent duplicate entries. Covering over 40,000 species, this dataset is designed to support tasks in biodiversity research and ecosystem dynamics studies.