# Paralinguistic-aware dialogue on ParaS2SBench Source: [Paralinguistic-aware dialogue on ParaS2SBench]() Collection index: [llms.txt]() Data updated: 2026-03-09T06:59:19.327Z Comprehensive evaluation of dialogue models on the ParaS2SBench dataset, covering paralinguistic awareness across both synthetic (Age, Emotion, Gender, Sarcasm) and real-world (IEMOCAP, MELD) components. A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Paralinguistic-aware dialogue | | Dataset (including any stated split) | ParaS2SBench | | Metric | Synthetic Age Score | | Unit | SCALAR | | Direction | Higher is better | | Results loaded | 11 of 11 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [Synthetic Age Score](): Higher is better; unit: SCALAR. - [Synthetic Emotion Score](): Higher is better; unit: SCALAR. - [Synthetic Gender Score](): Higher is better; unit: SCALAR. - [Synthetic Sarcasm Score](): Higher is better; unit: SCALAR. - [Synthetic Average Score](): Higher is better; unit: SCALAR. - [Real IEMOCAP Score](): Higher is better; unit: SCALAR. - [Real MELD Score](): Higher is better; unit: SCALAR. - [Real Average Score](): Higher is better; unit: SCALAR. - [Overall Average Score](): Higher is better; unit: SCALAR. ## Recorded results | Method | Synthetic Age Score | Result date | Report paper | | --- | --- | --- | --- | | GPT-TTS | 4.525 | Not recorded | 691566140052cac1dc65d8c4 | | Kimi-Audio GRPO | 4.496 | Not recorded | 691566140052cac1dc65d8c4 | | Kimi-Audio SFT | 4.393 | Not recorded | 691566140052cac1dc65d8c4 | | Gemini | 3.301 | Not recorded | 691566140052cac1dc65d8c4 | | gpt-4o-audio-preview | 3.205 | Not recorded | 691566140052cac1dc65d8c4 | | Qwen2.5 Omni | 3.17 | Not recorded | 691566140052cac1dc65d8c4 | | Kimi-Audio | 3.141 | Not recorded | 691566140052cac1dc65d8c4 | | LLaMa-Omni 2 | 3.123 | Not recorded | 691566140052cac1dc65d8c4 | | Whisper-GPT-TTS | 3.05 | Not recorded | 691566140052cac1dc65d8c4 | | GLM 4 | 2.885 | Not recorded | 691566140052cac1dc65d8c4 | | Freeze-Omni | 2.819 | Not recorded | 691566140052cac1dc65d8c4 | ### GPT-TTS Synthetic Age Score: 4.525. Result ID: 69ae6e916dbb7075b955f650. [Reporting paper]() Recorded evaluation configuration: {"Category":"Topline"}. This record does not assert independent reproduction. ### Kimi-Audio GRPO Synthetic Age Score: 4.496. Result ID: 69ae6e916dbb7075b955f64f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Category":"Ours"}. This record does not assert independent reproduction. ### Kimi-Audio SFT Synthetic Age Score: 4.393. Result ID: 69ae6e916dbb7075b955f64e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Category":"Ours"}. This record does not assert independent reproduction. ### Gemini Synthetic Age Score: 3.301. Result ID: 69ae6e916dbb7075b955f648. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Category":"Closed Source"}. This record does not assert independent reproduction. ### gpt-4o-audio-preview Synthetic Age Score: 3.205. Result ID: 69ae6e916dbb7075b955f647. [Reporting paper]() Recorded evaluation configuration: {"Category":"Closed Source"}. This record does not assert independent reproduction. ### Qwen2.5 Omni Synthetic Age Score: 3.17. Result ID: 69ae6e916dbb7075b955f649. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Category":"Open Source"}. This record does not assert independent reproduction. ### Kimi-Audio Synthetic Age Score: 3.141. Result ID: 69ae6e916dbb7075b955f64d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Category":"Open Source"}. This record does not assert independent reproduction. ### LLaMa-Omni 2 Synthetic Age Score: 3.123. Result ID: 69ae6e916dbb7075b955f64b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Category":"Open Source"}. This record does not assert independent reproduction. ### Whisper-GPT-TTS Synthetic Age Score: 3.05. Result ID: 69ae6e916dbb7075b955f646. [Reporting paper]() Recorded evaluation configuration: {"Category":"Baseline"}. This record does not assert independent reproduction. ### GLM 4 Synthetic Age Score: 2.885. Result ID: 69ae6e916dbb7075b955f64a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Category":"Open Source"}. This record does not assert independent reproduction. ### Freeze-Omni Synthetic Age Score: 2.819. Result ID: 69ae6e916dbb7075b955f64c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Category":"Open Source"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.