โ† All examples

ModelEthio-ASR Amharic (600M)5Ethio-ASR Multi (300M)10Ethio-ASR Multi (600M)9Ethio-ASR Multi (94M)8Meta MMS-1B12OmniASR-CTC (1B)1OmniASR-CTC (300M)3OmniASR-CTC (3B)0OmniASR-LLM (1B)1OmniASR-LLM (300M)1OmniASR-LLM (3B)1OpenAI Whisper-small12
Audio qualityClean Studio3Ambient Room Noise4Harsh Environment4Distant / Low-Volume Mic3Wideband Phone (VoLTE)4Randomized Phone Channel8Narrowband 2G + Packet Loss10

What did the model get wrong?

Ethio-ASR Multi (94M) Randomized Phone Channel   Example 20 of 20  ยท  speaker: Male

๐ŸŽง Listen โ€” Randomized Phone Channel
Phone-call audio with random extra distortions layered on top.
โœ… What was actually said
แ‰แŒฅแˆฉ แ‹œแˆฎ แ‹˜แŒ แŠ แˆแˆˆแ‰ต แ‹œแˆฎ แŠ แˆแˆตแ‰ต แˆตแ‹ตแˆตแ‰ต แˆฐแ‰ฃแ‰ต แˆตแˆแŠ•แ‰ต แ‹˜แŒ แŠ แ‹œแˆฎ แŠแ‹
๐Ÿค– What this model heard
amh แ‰แŒฅแˆฉ แˆฎ แ‹˜แŒฃแŠ แˆแˆˆแ‰ต แ‹œแˆฎ แŠ แˆแˆตแ‰ตแ‹ตแˆตแ‰ต แˆฐแ‰ฃแŠ แˆแŠ•แ‰ต แ‹˜แŒฃแŠ แ‹œแˆฎ แŠแ‹
Correct
Wrong (different word)
Missed (skipped)
Extra (added)
5
Correct
6
Wrong
1
Missed
1
Extra
42%
Words right
Out of 12 spoken words, this model got 5 right (42%). Errors: 6 words wrong; 1 missed; 1 extra.

Try switching Audio quality โ€” see the same model degrade as the audio gets noisier.