Test splits for CREMA-D, emoDB, IEMOCAP, MELD, RAVDESS

Test splits for the categorical emotion datasets CREMA-D, emoDB, IEMOCAP, MELD, RAVDESS used inside audEERING. 
For each dataset, a CSV file is provided listing the file names included in the test split.
The test splits were designed trying to balance gender and emotional categories as good as possible.

本数据集为audEERING内部使用的分类情感数据集CREMA-D、emoDB、IEMOCAP、MELD、RAVDESS的测试划分集。针对每个数据集，均提供一份CSV文件，其中列出了该测试划分集所包含的文件名。本次测试划分集的设计力求尽可能均衡性别与情感类别的分布。