Tarteel AI's EveryAyah dataset contains Quranic verse audio recordings and their corresponding transcriptions. The collection features multiple reciters and includes diacritization in the text.
Use Cases
- Train an automatic speech recognition model on the Arabic audio and text transcription pairs.
- Analyze diacritization patterns across different reciters using the provided text transcriptions.
- Build a speaker identification or reciter classification model from the collection of audio files.
Strengths
- Data includes diacritization in transcriptions, which is important for Arabic text analysis.
- Features audio from multiple reciters, allowing for comparative analysis.
Limitations
- Dataset size, row count, and specific audio duration statistics are unknown.
- The dataset's temporal coverage and geographic origin of recitations are not specified.
Provenance
- Source
- tarteel-ai
- Collection Method
- Collection of Quranic verses and their transcriptions by different reciters.
- Time Range
- null
- Freshness
- null
- Geography
- null