@jkh As for diacritics, the Riva Arabic ASR model supports diacritics but not every speech is transcribed with full diacritics, it depends on context. If the audio is in Modern Standard Arabic or dialectal speech then the model will provide only partial diacritics where the context is ambiguous or diacritics will aid meaning such as shaddah or tanween. If the speech is Quran, then the model will produce fully diacritized transcripts with harkat. So, the answer is yes diacritics are supported as a feature by the model but it’s context dependent.
Related topics
Topic | Replies | Views | Activity | |
---|---|---|---|---|
Riva on Whisper Large v3 returns only part transcription | 7 | 115 | January 23, 2025 | |
Canary 1b producing 'x's as transcription on Arabic audio | 5 | 41 | January 23, 2025 | |
Arabic ASR using riva throws error - "Error: Unavailable model requested given these parameters: language_code=ar; sample_rate=16000; type=offline; " | 0 | 45 | February 25, 2025 | |
Issue Deploying Fine-Tuned Arabic Conformer Model in NVIDIA Riva: No Transcriptions Returned | 0 | 69 | December 1, 2024 | |
Would like to know the ASR model used in demo page | 1 | 83 | July 1, 2024 | |
Riva ASR not returning the final transcipt accurately | 1 | 56 | June 9, 2025 | |
Finetuned ASR conformer returns only empty transcripts | 13 | 996 | October 20, 2022 | |
List ASR models support by nemo2riva conversion | 5 | 113 | May 14, 2025 | |
Where can I try out ASR for Arabic? | 1 | 24 | November 30, 2024 | |
Deploying NVIDIA Riva Multilingual ASR with Whisper and Canary Architectures While Selectively Deactivating NMT | 1 | 51 | February 20, 2025 |