| --- |
| license: mit |
| --- |
| |
| # RAVDESS |
|
|
| This is an audio classification dataset for **Emotion Recognition**. |
|
|
| **Classes = 8 , Split = Train-Test** |
|
|
|
|
| ## Structure |
| - `audios` folder contains audio files. |
| - `train.csv` for training split and `test.csv` for the testing split. |
|
|
| ## Download |
| ```python |
| import os |
| import huggingface_hub |
| audio_datasets_path = "DATASET_PATH/Audio-Datasets" |
| if not os.path.exists(audio_datasets_path): print(f"Given {audio_datasets_path=} does not exist. Specify a valid path ending with 'Audio-Datasets' folder.") |
| huggingface_hub.snapshot_download(repo_id="MahiA/RAVDESS", repo_type="dataset", local_dir=os.path.join(audio_datasets_path, "RAVDESS")) |
| ``` |
|
|
|
|
| ## Acknowledgment |
| This dataset is a slightly processed/restructured version of data originally released by [Source](https://zenodo.org/records/1188976#.YFZuJ0j7SL8).<br> |
| Please refer to the respective source for their licensing details and any additional information. |
|
|
|
|
|
|
| ## Contact |
| For questions or feedback, please create an issue. |
|
|
|
|
