This dataset is the audio subset of the **ShEMO** Persian Emotional Speech Dataset. In this version, files labeled with the emotion fear have been removed due to the limited number of such instances (only 38 files in total), making them unsuitable for reliable experimentation in this research. The dataset is based on the **modified version of ShEMO**, which can be accessed at this [repository](https://github.com/aliyzd95/modified_shemo).
