| Property | Value |
|---|---|
| Author | Nuwaisir |
| Framework | PyTorch |
| Base Architecture | wav2vec2 |
| Downloads | 85 |
What is Quran_speech_recognizer?
Quran_speech_recognizer is an innovative speech recognition model specifically designed for Quranic recitation. It utilizes transfer learning techniques by fine-tuning the wav2vec2-large-xlsr-53-arabic pre-trained model to accurately recognize and locate Quranic verses from spoken recitation.
Implementation Details
The model leverages transfer learning by fine-tuning a pre-existing Arabic speech recognition model using the Quran ASR Challenge dataset. It implements real-time speech processing capabilities and can process 5-second audio segments to identify corresponding Quranic verses, currently optimized for the 30th juzz (Surah 78-114).
- Built on wav2vec2 architecture for robust speech recognition
- Fine-tuned using specialized Quranic recitation dataset
- Implements edit distance algorithm for verse matching
- Real-time audio processing capabilities
Core Capabilities
- Recognition of Quranic recitations in real-time
- Automatic verse position identification
- Support for 30th juzz of the Quran (extensible)
- Interactive user interface through Jupyter notebook
Frequently Asked Questions
Q: What makes this model unique?
This model specifically targets Quranic recitation recognition, utilizing transfer learning from a robust Arabic speech recognition model and specialized fine-tuning for Quranic verses.
Q: What are the recommended use cases?
The model is ideal for educational settings, self-study of Quran recitation, and automatic identification of Quranic verses from audio input. It's particularly useful for students and teachers of Quranic studies.