Lexical Tone is Hard to Quantize: Probing Discrete Speech Units in Mandarin and Yor\`ub\'a

Build a diagnostic tool to evaluate how well current quantization schemes preserve lexical tone in speech units. Provide a 'tone-preservation' score for researchers building TTS in tonal languages.

Suggested repo: toneProbe

"Don't lose your tone in quantization."

Estimated effort: 40h