Acoustic data-driven lexicon learning based on a greedy pronunciation selection framework

Re-assign community