Egyptian and Nigerian English join the Open ASR Leaderboard

The Open ASR Leaderboard has added private conversational English evaluation data for Egyptian and Nigerian accents through a collaboration with Liva AI, contributing 1.34 and 1.55 hours of spontaneous, punctuated, and disfluency-inclusive speech respectively. The additions extend earlier private benchmark splits from Appen, DataOceanAI, and Voice Arena, which were introduced to improve evaluation reliability and reduce optimization specifically targeted at public benchmarks. The leaderboard will also provide expanded accent-level reporting based on Kachru’s Three Circles framework, comparing US and non-US English as well as Inner Circle accents with Outer and Expanding Circle accents. Liva AI supplies naturally occurring multimodal interaction data collected through its consumer products, enabling datasets designed around gaps in model performance across languages, domains, and interaction types.

Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.