---
license: openrail
dataset_info:
  features:
  - name: audio
    dtype: audio
  - name: sentence
    dtype: string
  splits:
  - name: train
    num_bytes: 7786139475.04
    num_examples: 17688
  - name: test
    num_bytes: 1447570290.631
    num_examples: 3317
  download_size: 8857354186
  dataset_size: 9233709765.671
---

# Dataset Card for "myst_pfs" 

Contains MyST dataset and PFstar datset as mentioned in our paper:

R. Jain, A. Barcovschi, M. Y. Yiwere, P. Corcoran and H. Cucu, "Exploring Native and Non-Native English Child Speech Recognition With Whisper," in IEEE Access, vol. 12, pp. 41601-41610, 2024, doi: 10.1109/ACCESS.2024.3378738.


### Dataset Card for `myst_wav_clean`

This is a subset of the MyST dataset used in our research experiments. It is distributed under the original MyST dataset license, available here: [LDC2021S05 – MyST Corpus](https://catalog.ldc.upenn.edu/LDC2021S05).

If you would like access to this subset, please provide proof of a valid license for the original dataset from the LDC.


### Dataset Card for `pfstar_subset`

This is a subset of the PFSTAR dataset used in our research experiments. It is distributed under the original PFSTAR dataset license. For licensing and access details, please refer to the following resources:

- 🔗 [ELRA Catalogue – PFSTAR Speech Corpus](http://universal.elra.info/product_info.php?cPath=37_39&products_id=1939)  
- 📄 [Batliner et al., Interspeech 2005 – PFSTAR Overview](https://www.isca-archive.org/interspeech_2005/batliner05b_interspeech.pdf)

If you would like access to this subset, please provide proof of a valid license or institutional access to the original PFSTAR dataset.


**Update (September 18, 2025):**  
We’ve received numerous requests for access to this dataset. If you're submitting a request via Hugging Face, please also send an email to **j_rishabh@outlook.com**. Make sure to include documentation confirming your access rights to the original PFSTAR dataset. This will help us verify and process your request more efficiently.