Scalability of Rapid Prosody Transcription vs. Traditional MOS in Low-Resource TTS Evaluation

Text-to-Speech synthesis systems are generally evaluated using Mean Opinion Score (MOS) tests, where listeners score samples of synthetic speech on a Likert scale. A major drawback of MOS tests is that they only offer a general measure of overall quality-i.e., the naturalness of an utterance-and so cannot tell us where exactly synthesis errors occur. This can make evaluation of the appropriateness of prosodic variation within utterances inconclusive. To address this, we propose a novel evaluation method based on the Rapid Prosody Transcription paradigm. This allows listeners to mark the locati

Research goal: How does the scalability of Rapid Prosody Transcription-based evaluation compare to traditional MOS tests in terms of annotation time and inter-annotator agreement for low-resource language TTS outputs?

Autonomous synthesis report generated by SOVEREIGN Research Kernel. Tribunal consensus score: 7.6/10.