List your company
ResearchFormerly WizWand

Automatic Speech Recognition on Low-resource Dialectal Speech 5 languages (test)

28.6WERCLD
23.935655.420386.905118.3897May 22, 2026
Updated 5mo ago

Evaluation results

MethodLinks
CLDASR Model Backbone=Whisper-Large-v3 (WSP-L)2026.0528.615.37
NNASR Model Backbone=Whisper-Large-v3 (WSP-L)2026.0529.2515.99
CLDASR Model Backbone=Whisper-Small (WSP)2026.0531.7417.84
Kernel SVMASR Model Backbone=Whisper-Large-v3 (WSP-L)2026.0537.9122.05
Linear SVMASR Model Backbone=Whisper-Large-v3 (WSP-L)2026.0539.3623.68
DefaultASR Model Backbone=Whisper-Large-v3 (WSP-L)2026.0540.4121.8
KNNASR Model Backbone=Whisper-Large-v3 (WSP-L)2026.0544.8929.12
CLDASR Model Backbone=MMS-1B2026.0545.2721.58
Kernel SVMASR Model Backbone=Whisper-Small (WSP)2026.0546.5226.14
NNASR Model Backbone=MMS-1B2026.0548.2623.64
Linear SVMASR Model Backbone=Whisper-Small (WSP)2026.0548.7428.28
Kernel SVMASR Model Backbone=MMS-1B2026.0549.1225.88
Linear SVMASR Model Backbone=MMS-1B2026.0550.7326.07
DefaultASR Model Backbone=MMS-1B2026.0551.8827.61
NNASR Model Backbone=Whisper-Small (WSP)2026.0553.8434.52
KNNASR Model Backbone=MMS-1B2026.0557.3432.76
DefaultASR Model Backbone=Whisper-Small (WSP)2026.05139.3773.85
KNNASR Model Backbone=Whisper-Small (WSP)2026.05145.2181.05