zaibihassan/Quranic-Word-By-Word-Audio-Data

《古兰经》逐词音频数据集（Muallim 和 Mujawwad）是一个包含《古兰经》两种完整逐词诵读风格的数据集，专为边缘交付、移动流媒体和机器学习管道优化。数据集包括：1. Muallim（教师风格）：针对慢速、教育和可重复聆听优化；2. Mujawwad（塔吉威德风格）：针对具有完整塔吉威德流程的自然节奏诵读优化。原始音频大小约为2.0-2.3GB，通过优化的OPUS工作流重新编码，压缩至约400MB，同时保持相同的听觉质量和时间精度。数据集包含114个苏拉（章节），超过77,000个独立的单词音频文件，采用严格的SURAH_AYAH_WORD索引系统命名。音频格式为16kHz单声道OPUS流，启用可变比特率（VBR）和语音优化调谐。每个单词音频段都包含同步的时间戳元数据，使用紧凑的Protocol Buffer (.pb)文件，支持实时高亮、卡拉OK式播放、翻译同步、AI语音对齐、波形搜索和多诵读者同步。数据集适用于自动语音识别、文本到语音和音频分类等任务，语言为阿拉伯语，许可证为Apache 2.0。

The Quran Word-By-Word Audio Dataset (Muallim & Mujawwad) contains two complete word-by-word recitation datasets of the Holy Quran, optimized for edge delivery, mobile streaming, and machine learning pipelines. The dataset includes: 1. Muallim (Teacher Style) — optimized for slow, educational, and repeat-friendly listening; 2. Mujawwad (Tajweed Style) — optimized for natural rhythmic recitation with full tajweed flow. Originally averaging between 2.0 GB to 2.3 GB each in raw format, the audio has been re-encoded using an optimized OPUS workflow, reduced to approximately 400 MB while preserving identical audible quality and timing precision. The dataset covers all 114 Surahs of the Holy Quran, with over 77,000 individual word audio files mapped using a strict SURAH_AYAH_WORD indexing system. Audio is compressed into 16kHz mono .opus streams with Variable Bitrate (VBR) enabled and speech-optimized tuning. Each word audio segment includes synchronized timestamp metadata using compact Protocol Buffer (.pb) files, enabling real-time highlighting, karaoke-style playback, translation synchronization, AI speech alignment, waveform seeking, and multi-reciter synchronization. It is suitable for tasks such as automatic speech recognition, text-to-speech, and audio classification, with language in Arabic and licensed under Apache 2.0.