Sada-ASR
A production-grade speech recognition system for Arabic dialects, developed by EchoNova, intended to address the performance degradation of Modern Standard Arabic recognizers in dialect scenarios.
Timeline (last 90 days)
Sada-ASR is a production-grade Arabic voice assistant/speech recognition system developed by EchoNova.
The Sada-ASR team built a 120-hour test set covering MSA and eight Arabic dialects.
On the 120-hour test set, Sada-ASR achieved an overall word error rate (WER) of 11.69%, while the strongest comparison system, Microsoft MAI-Transcribe-2, achieved 13.51%.
On segments unaffected by digits and writing conventions, Sada-ASR's WER lead over Microsoft MAI-Transcribe-2 was 0.74 percentage points.
Sada-ASR's average word error rate across dialects was 15.52%, compared with an average of 16.96% for the other comparison systems.
Sada-ASR leads competitors by 4.30 percentage points on Omani dialect, 3.78 percentage points on Egyptian dialect, and 2.19 percentage points on Saudi dialect.
On Lebanese dialect, Sada-ASR's word error rate was 24.96%, trailing the competitor's 23.05%.
The Sada-ASR team explains that the main reason for the weaker performance on Lebanese dialect is that it transcribes Lebanese vocabulary into MSA forms, resulting in lower verbatim transcription accuracy.
The Sada-ASR team open-sourced a 20-hour subset of the test set along with the scoring code.
All Facts (9)
Sada-ASR is a production-grade Arabic voice assistant/speech recognition system developed by EchoNova.
50%UnverifiedThe Sada-ASR team built a 120-hour test set covering MSA and eight Arabic dialects.
50%UnverifiedOn the 120-hour test set, Sada-ASR achieved an overall word error rate (WER) of 11.69%, while the strongest comparison system, Microsoft MAI-Transcribe-2, achieved 13.51%.
50%UnverifiedOn segments unaffected by digits and writing conventions, Sada-ASR's WER lead over Microsoft MAI-Transcribe-2 was 0.74 percentage points.
50%UnverifiedSada-ASR's average word error rate across dialects was 15.52%, compared with an average of 16.96% for the other comparison systems.
50%UnverifiedSada-ASR leads competitors by 4.30 percentage points on Omani dialect, 3.78 percentage points on Egyptian dialect, and 2.19 percentage points on Saudi dialect.
50%UnverifiedOn Lebanese dialect, Sada-ASR's word error rate was 24.96%, trailing the competitor's 23.05%.
50%UnverifiedThe Sada-ASR team explains that the main reason for the weaker performance on Lebanese dialect is that it transcribes Lebanese vocabulary into MSA forms, resulting in lower verbatim transcription accuracy.
50%UnverifiedThe Sada-ASR team open-sourced a 20-hour subset of the test set along with the scoring code.
50%