Egyptian Arabic Speech Dataset with Emotion Labels and Speaker Diarization

An Egyptian Arabic dataset combining text and audio, annotated with emotions and speaker diarization. Created by OmarAhmedSobhy, it is designed for training Text-to-Speech and Automatic Speech Recognition models. The dataset was last updated on May 7, 2026.

Use Cases

• Training expressive Text-to-Speech models based on the described emotion labels.
• Developing Automatic Speech Recognition systems for Egyptian Arabic based on the paired audio-text data.
• Researching speaker diarization techniques based on the included speaker_role metadata.
• Filtering data for single-speaker experiments based on the speaker_role field mentioned in the description.

Strengths

• Includes emotion labels, which can enhance model expressiveness.
• Contains speaker diarization information, enabling multi-speaker analysis.
• Specifically targets the Egyptian Arabic dialect, filling a potential niche.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Data may reflect geographic and source bias inherent to its collection method from platforms like YouTube.

Provenance

Source

Likely collected from YouTube, as indicated by the source code link.

Collection Method

Created using forced alignment and ASR models (MahmoudAshraf/mms-300m-1130-forced-aligner, Cohere ASR, Hubert).

Time Range

null

Freshness

Last updated 2026-05-07 13:24:26; freshness should be verified.

Geography

Egyptian Arabic dialect.