All datasets / obadx

obadx/recitation-segmentation

225 downloads in 30 days1 likesLast modified 3 Sep 2025View on Hugging Face

Permissive or share-alike

The declared licence (mit) is classified as permissive.

Declared licence: mit, from the tagEvidence: README.md at bec90c4
Licence files: No licence file at the top level of the repositoryEvidence: file list at bec90c4
Declared sources: None listed in the card metadata
Access gated on Hugging Face: No

Each cell links to what was checked: a file, a file list or a Hub record.

Tasks

automatic-speech-recognition

Citation text from the card

This dataset is used for building models that segment Holy Quran recitations based on pause points (waqf) with high accuracy. The segments are crucial for tasks like Automatic Pronunciation Error Detection and Correction, leveraging the rigorous recitation rules (tajweed) of the Holy Quran.

The dataset was presented in the paper [Automatic Pronunciation Error Detection and Correction of the Holy Quran's Learners Using Deep Learning](https://huggingface.co/papers/2509.00094).

- **Project Page**: [https://obadx.github.io/prepare-quran-dataset/](https://obadx.github.io/prepare-quran-dataset/)
- **Code/Model Repository**: [https://github.com/obadx/recitations-segmenter](https://github.com/obadx/recitations-segmenter)

This dataset comprises **850+ hours of audio (~300K annotated utterances)** and was generated through a 98% automated pipeline, which includes:
*   Collection of recitations from expert reciters.
*   Segmentation at pause points (waqf) using a fine-tuned wav2vec2-BERT model.
*   Transcription of segments.
*   Transcript verification via the novel Tasmeea algorithm.

A novel Quran Phonetic Script (QPS) is employed to encode Tajweed rules, distinguishing it from standard IPA for Modern Standard Arabic. This high-quality annotated data addresses the scarcity of resources for Quranic speech processing and is crucial for developing robust ASR-based approaches for pronunciation error detection and correction.

All datasets marked Permissive or share-alike | JSON

The same facts for your agent

A free MCP server at mcp.datalicense.uk/mcp. Connect it from an MCP client such as Claude or Cursor.

dataset_licence({ "id": "obadx/recitation-segmentation" })
signal: "permissive"
stated_licence: "mit"
sources: 0