Figure 4From: Optimizing Automatic Speech Recognition for Low-Proficient Non-Native Speakers Distribution of lexicon sizes. Back to article page