# Multi-Dimensional Prosody Judgment For Live Streaming Speech Synthesis

arXiv: https://arxiv.org/abs/2609.20124 · HTML: https://mlsanity.com/paper/2609.20124/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 18 of its 19 references that are in the map.
- First posted on arXiv 17 Sep 2026.

## Builds on (most cited first, 18 of 18)

- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [Qwen3-Omni Technical Report](https://mlsanity.com/paper/2509.17765/) (2025)
- [UTMOS: UTokyo-SaruLab System for VoiceMOS Challenge 2022](https://mlsanity.com/paper/2204.02152/) (2022)
- [SWIFT:A Scalable lightWeight Infrastructure for Fine-Tuning](https://mlsanity.com/paper/2408.05517/) (2024)
- [CosyVoice: A Scalable Multilingual Zero-shot Text-to-speech Synthesizer based on Supervised Semantic Tokens](https://mlsanity.com/paper/2407.05407/) (2024)
- [NISQA: A Deep CNN-Self-Attention Model for Multidimensional Speech Quality Prediction with Crowdsourced Datasets](https://mlsanity.com/paper/2104.09494/) (2021)
- [MOSNet: Deep Learning based Objective Assessment for Voice Conversion](https://mlsanity.com/paper/1904.08352/) (2019)
- [Qwen3-TTS Technical Report](https://mlsanity.com/paper/2601.15621/) (2026)
- [IndexTTS2: A Breakthrough in Emotionally Expressive and Duration-Controlled Auto-Regressive Zero-Shot Text-to-Speech](https://mlsanity.com/paper/2506.21619/) (2025)
- [FireRedTTS: A Foundation Text-To-Speech Framework for Industry-Level Generative Speech Applications](https://mlsanity.com/paper/2409.03283/) (2024)
- [EmergentTTS-Eval: Evaluating TTS Models on Complex Prosodic, Expressiveness, and Linguistic Challenges Using Model-as-a-Judge](https://mlsanity.com/paper/2505.23009/) (2025)
- [SpeechJudge: Towards Human-Level Judgment for Speech Naturalness](https://mlsanity.com/paper/2511.07931/) (2025)
- [GSRM: Generative Speech Reward Model for Speech RLHF](https://mlsanity.com/paper/2602.13891/) (2026)
- [dots.tts Technical Report](https://mlsanity.com/paper/2606.07080/) (2026)
- [Evaluating the Expressive Appropriateness of Speech in Rich Contexts](https://mlsanity.com/paper/2605.09413/) (2026)
- [UniSRM: A Unified Speech Reward Model for Reasoning-Based Fine-grained Assessment](https://mlsanity.com/paper/2605.23261/) (2026)
- [Comprehensive Benchmarking of Long-Form Speech Generation in Diverse Scenarios](https://mlsanity.com/paper/2605.28618/) (2026)

## Abstract (arXiv)

Evaluating live streaming speech synthesis (TTS) requires assessing fine-grained, highly expressive prosody such as emotion, intonation, and energy which traditional MOS predictors fail to capture. While proprietary Large Language Models (LLMs) like Gemini can evaluate these aspects, they are too costly for massive inference and reinforcement learning feedback. To address this, we first introduce Live-ProsodyJudge (LPJ), a cost-effective pairwise evaluator distilled from Gemini into Qwen3-Omni. However, we identify a critical flaw in standard multi-dimensional evaluation: verdict coupling. The judge tends to lazily align all individual dimension scores with its overall preference, collapsing a rich multi-dimensional rubric into a single preference bit. To resolve this, we further propose Decoupled-Live-ProsodyJudge (D-LPJ). D-LPJ eliminates the overall verdict target to prevent blind following, masks uncertain pair-dimensions during Supervised Fine-Tuning(SFT), and introduces a novel span-local GRPO strategy that applies normalized advantages strictly to their corresponding rationale spans. Evaluated on highly curated human-annotated test sets, 10 sample balanced-order LPJ achieves higher point accuracy than a single Gemini call, while D-LPJ successfully produces independent,decoupled dimension judgments. Furthermore, in a Best-of-8 TTS candidate selection tournament, the LPJ-selected utterance falls within the human top-3 in 85.29% of high-confidence cases, demonstrating its efficacy for fine-grained TTS preference optimization.