Fact · Languages and accents
The 2017 MGB-3 Arabic challenge emphasized dialectal Arabic and drew its data from Egyptian YouTube videos.
Checked against its source on October 17, 2026
In the source's own words
“MGB-3 emphasises dialectal Arabic using a multi-genre collection of Egyptian YouTube videos”
Speech Recognition Challenge in the Wild: Arabic MGB-3, Ahmed Ali et al., arXiv, 2017, Abstract.
Open the original- What it measures
- The challenge moved away from the earlier broadcast news test (from Al Jazeera) to casual, dialectal speech, which is harder for systems trained on news.
- What it does not show
- A research challenge dataset, not a product test. It shows what a dialect challenge looks like, not how a dictation app performs.